[
  {
    "category": "Image",
    "description": "General-purpose workflow templates for native image generation, including text-to-image, image editing, inpainting, and other core image workflows supported out of the box.",
    "templates": [
      {
        "name": "api_nano_banana_2_1_t2i",
        "title": "Nano Banana 2.1: Text to Image",
        "task": "Text to Image",
        "model": "Nano Banana",
        "freshness": "new",
        "usage": 5111,
        "recommend": "highly_recommended",
        "description": "Generate images from text prompts using Nano Banana 2.1, Google's fast Gemini-based image model, delivering one finished image with no input files required. Ideal for brand marketing posters, e-commerce product visuals, and IP character series.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.39.1",
        "capabilities": {
          "workflow": [
            "text-to-image",
            "api"
          ],
          "model_options": {
            "GeminiNanoBanana2V2": [
              "Gemini Nano Banana 2.1",
              "Nano Banana 2 (Gemini 3.1 Flash Image)",
              "Nano Banana 2 Lite"
            ]
          }
        }
      },
      {
        "name": "image_z_image_turbo",
        "title": "Z-Image-Turbo: Text to Image",
        "task": "Text to Image",
        "model": "Z-Image-Turbo",
        "freshness": "established",
        "usage": 14712,
        "recommend": "highly_recommended",
        "description": "An Efficient Image Generation Foundation Model with Single-Stream Diffusion Transformer, supports English & Chinese.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.11.0",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "image_qwen_image_2_1_image_edit",
        "title": "Qwen Image 2.1: Image Edit",
        "task": "Image Edit",
        "model": "Qwen Image 2.1",
        "freshness": "recent",
        "usage": 15631,
        "recommend": "highly_recommended",
        "description": "Edit images with Qwen-Image-2.1 by feeding two reference images through a single checkpoint that handles both generation and editing, then comparing the result against the original. Ideal for character-consistent edits, product cutout compositing with real transparency, and style-transfer passes at native 2K.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.37.0",
        "capabilities": {
          "workflow": [
            "image-edit",
            "int8"
          ]
        }
      },
      {
        "name": "image_p_h3_image_edit",
        "title": "P-H3-Image: Image Edit",
        "task": "Image Edit",
        "model": "MiniMax H3",
        "freshness": "new",
        "usage": 269,
        "recommend": "medium",
        "description": "Pruna AI's few-step image model built on MiniMax-H3: edit the source image with a short instruction, up to 9 reference images, 8-step sampling.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.44.0",
        "capabilities": {
          "workflow": [
            "image-edit",
            "int8"
          ]
        }
      },
      {
        "name": "image_p_h3_t2i",
        "title": "P-H3-Image: Text to Image",
        "task": "Text to Image",
        "model": "MiniMax H3",
        "freshness": "new",
        "usage": 245,
        "recommend": "medium",
        "description": "Pruna AI's few-step image model built on MiniMax-H3: write a plain-language request and a local prompt upsampler expands it into a detailed scene before an 8-step sample. Pick the aspect ratio and megapixel target on the Resolution Selector to move between square, 4:3, 16:9 and portrait canvases.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.44.0",
        "capabilities": {
          "workflow": [
            "text-to-image",
            "int8"
          ]
        }
      },
      {
        "name": "image_qwen_image_2_1_t2i",
        "title": "Qwen Image 2.1: Text to Image",
        "task": "Text to Image",
        "model": "Qwen Image 2.1",
        "freshness": "recent",
        "usage": 2457,
        "recommend": "top",
        "description": "Generate images from text prompts with Qwen Image 2.1, the latest open-weight Qwen-Image model, producing a single 1024×1024 image over 25 sampling steps. Generation and instruction-based editing share one model with native 2K output, professional typography, and alpha-channel support. Ideal for poster and layout design, transparent asset creation, and rapid concept visualization.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.37.2",
        "capabilities": {
          "workflow": [
            "text-to-image",
            "int8"
          ]
        }
      },
      {
        "name": "image_krea2_turbo_t2i",
        "title": "Krea-2: Text to Image",
        "task": "Text to Image",
        "model": "Krea 2",
        "freshness": "current",
        "usage": 9480,
        "recommend": "high",
        "description": "Generate images from text prompts using Krea 2, a foundation model built for aesthetic quality and creative control. It focuses on rendering expressive and artistic styles. Ideal for concept art exploration, visual brainstorming, and creating stylized imagery for design projects.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.24.1",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "image_qwen_image_edit_2509",
        "title": "Qwen Image Edit 2509",
        "task": "Image Edit",
        "model": "Qwen-Image",
        "freshness": "established",
        "usage": 3470,
        "recommend": "highly_recommended",
        "description": "Advanced image editing with multi-image support, improved consistency, and ControlNet integration.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.23.0",
        "capabilities": {
          "workflow": [
            "controlnet",
            "image-edit"
          ]
        }
      },
      {
        "name": "image_qwen_image_2_1_turbo_image_edit",
        "title": "Qwen Image 2.1 Turbo: Image edit",
        "task": "Image Edit",
        "model": "Qwen Image 2.1",
        "freshness": "new",
        "usage": 2986,
        "recommend": "highly_recommended",
        "description": "Edit images with Qwen-Image-2.1 by feeding two reference images through a single checkpoint that handles both generation and editing, then comparing the result against the original. Ideal for character-consistent edits, product cutout compositing with real transparency, and style-transfer passes at native 2K",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.39.0",
        "capabilities": {
          "workflow": [
            "image-edit"
          ]
        }
      },
      {
        "name": "api_openai_gpt_image_26_sunburst_t2i",
        "title": "GPT Image 2.6 Sunburst: Text to Image",
        "task": "Text to Image",
        "model": "GPT Image 2.6",
        "freshness": "new",
        "usage": 0,
        "recommend": "medium",
        "description": "GPT Image 2.6 Sunburst is the detail-first half of GPT Image 2.6: it spends more of the render on fine texture, small type and materials, and keeps subjects and lighting consistent from one pass to the next. Use it for finished visuals, product shots and dense layouts where detail has to hold up.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.44.1",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-image"
          ],
          "model_options": {
            "OpenAIGPTImageNodeV2": [
              "gpt-image-2.5-flare",
              "gpt-image-2.5-sunburst",
              "gpt-image-2",
              "gpt-image-1.5",
              "gpt-image-1"
            ]
          }
        }
      },
      {
        "name": "image_qwen_image_2_1_turbo_t2i",
        "title": "Qwen Image 2.1 Turbo: Text to Image",
        "task": "Text to Image",
        "model": "Qwen Image 2.1",
        "freshness": "new",
        "usage": 1167,
        "recommend": "top",
        "description": "Generate images from text prompts with Qwen Image 2.1 Turbo, the 8-step accelerated checkpoint, producing a single 1024×1024 image in 8 sampling steps. Same model family with native 2K output, professional typography, and alpha-channel support, plus an optional Qwen 3.5 prompt enhancer. Ideal for poster and layout design, transparent asset creation, and rapid concept visualization.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.39.0",
        "capabilities": {
          "workflow": [
            "int8",
            "text-to-image",
            "turbo"
          ]
        }
      },
      {
        "name": "api_openai_gpt_image_26_flare_image_edit",
        "title": "GPT Image 2.6 Flare: Image Edit",
        "task": "Image Edit",
        "model": "GPT Image 2.6",
        "freshness": "new",
        "usage": 0,
        "recommend": "medium",
        "description": "Edit one detail of a photo and leave the rest alone: keep the face, pose, framing and lighting, then change the outfit, hair or a single object. Built on GPT Image 2.6 Flare, so drafts come back in seconds and the identity holds well enough to iterate on.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.44.1",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ],
          "model_options": {
            "OpenAIGPTImageNodeV2": [
              "gpt-image-2.5-flare",
              "gpt-image-2.5-sunburst",
              "gpt-image-2",
              "gpt-image-1.5",
              "gpt-image-1"
            ]
          }
        }
      },
      {
        "name": "api_openai_gpt_image_26_flare_t2i",
        "title": "GPT Image 2.6 Flare: Text to Image",
        "task": "Text to Image",
        "model": "GPT Image 2.6",
        "freshness": "new",
        "usage": 0,
        "recommend": "medium",
        "description": "GPT Image 2.6 Flare is OpenAI's speed-first image model: it returns a draft in seconds and holds the layout together on dense posters and product shots. Use it for concept passes, mood boards and typography-heavy drafts, then switch to a higher quality setting for the final image.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.44.1",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-image"
          ],
          "model_options": {
            "OpenAIGPTImageNodeV2": [
              "gpt-image-2.5-flare",
              "gpt-image-2.5-sunburst",
              "gpt-image-2",
              "gpt-image-1.5",
              "gpt-image-1"
            ]
          }
        }
      },
      {
        "name": "api_openai_gpt_image_26_sunburst_image_edit",
        "title": "GPT Image 2.6 Sunburst: Image Edit",
        "task": "Image Edit",
        "model": "GPT Image 2.6",
        "freshness": "new",
        "usage": 0,
        "recommend": "medium",
        "description": "Edit one thing in a photo and keep the rest identical: the same camera angle, framing, subject position and light, with only the requested element changed. GPT Image 2.6 Sunburst renders fine texture and small type, so posters and dense scenes hold up better on the second or third pass.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.44.1",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ],
          "model_options": {
            "OpenAIGPTImageNodeV2": [
              "gpt-image-2.5-flare",
              "gpt-image-2.5-sunburst",
              "gpt-image-2",
              "gpt-image-1.5",
              "gpt-image-1"
            ]
          }
        }
      },
      {
        "name": "image_minimax_h3_t2i",
        "title": "MiniMax H3: Text to Image",
        "task": "Text to Image",
        "model": "MiniMax H3",
        "freshness": "new",
        "usage": 983,
        "recommend": "high",
        "description": "Generate a single still image from a text prompt with MiniMax H3, an omni-modal video model used here for its first-frame decode. No file inputs required: the graph samples an empty latent and decodes it through the H3 video VAE. Best suited to a single subject with a defined light source and a simple or blurred background.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.30.0",
        "capabilities": {
          "workflow": [
            "text-to-image",
            "int8"
          ]
        }
      },
      {
        "name": "image_moss_image_t2i",
        "title": "MOSS-Image: Text to Image",
        "task": "Text to Image",
        "model": "MOSS-Image",
        "freshness": "new",
        "usage": 6885,
        "recommend": "highly_recommended",
        "description": "Generate an image from a text prompt with MOSS-Image, an open-weights model that reads a full scene caption and returns one finished image. A prompt upsampler expands a short request into that caption first, so a one-line description is enough. Ideal for editorial photography, product shots, and concept art.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.41.3",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "image_moss_image_base_t2i",
        "title": "MOSS-Image Base: Text to Image",
        "task": "Text to Image",
        "model": "MOSS-Image",
        "freshness": "new",
        "usage": 2254,
        "recommend": "top",
        "description": "Generate an image from a text prompt with the MOSS-Image base checkpoint, which reads a full scene caption and returns one finished image. The prompt upsampler expands a short request into that caption first. Use it for fine-tuning experiments and comparisons with the instruction-tuned checkpoint.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.41.3",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "image_moss_image_t2i_bboxes",
        "title": "MOSS-Image: Layout-Guided Text to Image",
        "task": "Text to Image",
        "model": "MOSS-Image",
        "freshness": "new",
        "usage": 352,
        "recommend": "medium",
        "description": "Generate an image from a text prompt plus a layout drawn in Create Bounding Boxes. Each box carries its own caption, and text boxes also render their text, so MOSS-Image places every element where you put it. Useful for posters, packaging shots, and any composition where placement matters.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.41.3",
        "capabilities": {
          "workflow": [
            "layout",
            "text-to-image"
          ]
        }
      },
      {
        "name": "image_moss_image_base_t2i_bboxes",
        "title": "MOSS-Image Base: Layout-Guided Text to Image",
        "task": "Text to Image",
        "model": "MOSS-Image",
        "freshness": "new",
        "usage": 128,
        "recommend": "medium",
        "description": "Generate an image from a text prompt plus a layout drawn in Create Bounding Boxes, using the MOSS-Image base checkpoint. Each box carries its own caption and text boxes render their text. Useful for comparing the base and instruction-tuned checkpoints on the same layout.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.41.3",
        "capabilities": {
          "workflow": [
            "text-to-image",
            "layout"
          ]
        }
      },
      {
        "name": "image_moss_image_edit",
        "title": "MOSS-Image: Image Edit LoRA",
        "task": "Image Edit",
        "model": "MOSS-Image",
        "freshness": "new",
        "usage": 1703,
        "recommend": "top",
        "description": "Edit a photo with a short instruction using MOSS-Image and its edit LoRA: the instruction and the source image are encoded together, so the sampler redraws the same frame and only the described change is applied. Colour, material, outfit and background swaps keep the face, hair, pose and framing intact, and a Compare Images node shows the source next to the result.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.43.0",
        "capabilities": {
          "workflow": [
            "image-edit"
          ]
        }
      },
      {
        "name": "image_moss_image_color_depth_control",
        "title": "MOSS-Image: Color & Depth Control",
        "task": "Depth Estimation",
        "model": "MOSS-Image",
        "freshness": "new",
        "usage": 217,
        "recommend": "medium",
        "description": "Re-pose and re-style a reference photo with MOSS-Image: the graph estimates the reference's geometry with MoGe 3, then generates a new image that keeps the pose, framing and silhouette while the prompt decides the character, wardrobe, setting and light. A color/depth control adapter LoRA makes the model follow the estimated map, and the color map socket is there for region-painted layouts.",
        "io": {
          "inputs": [
            "image: Input image"
          ],
          "outputs": [
            "image: Depth map"
          ]
        },
        "minComfyUIVersion": "0.43.0",
        "capabilities": {
          "workflow": [
            "depth"
          ]
        }
      },
      {
        "name": "api_bria_fibo2_t2i",
        "title": "Bria: FIBO 2.0 Text to Image",
        "task": "Text to Image",
        "model": "BRIA",
        "freshness": "new",
        "usage": 163,
        "recommend": "medium",
        "description": "Generate an image from a text prompt with Bria FIBO 2.0, one model for both generation and editing. It is trained on structured VGL captions, so a detailed poster-style brief is followed closely, exact wording is rendered literally, and the node also returns that structured prompt so the same image can be rebuilt or refined from JSON. Ideal for editorial covers, posters and typography-led layouts.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.42.3",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-image"
          ]
        }
      },
      {
        "name": "image_flux3_klein_image_edit",
        "title": "FLUX 3 Image Klein: Image Edit",
        "task": "Image Edit",
        "model": "FLUX 3 Image",
        "freshness": "new",
        "usage": 1262,
        "recommend": "top",
        "description": "Edit an existing image with FLUX 3 Image Klein by supplying one input image and two text prompts, producing one refined output alongside a side-by-side comparison view. Ideal for targeted retouching, style adjustments, and detail corrections that preserve the original composition.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.41.0",
        "capabilities": {
          "workflow": [
            "image-edit"
          ]
        }
      },
      {
        "name": "image_flux3_klein_distilled_image_edit",
        "title": "FLUX 3 Image Klein Distilled: Image Edit",
        "task": "Image Edit",
        "model": "FLUX 3 Image",
        "freshness": "new",
        "usage": 614,
        "recommend": "high",
        "description": "Edit an existing image with FLUX 3 Image Klein by pairing one input image with one text prompt, producing a single edited result and a side-by-side comparison view. Ideal for targeted retouching, restyling a photo, or swapping details while keeping the rest of the scene intact.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.41.0",
        "capabilities": {
          "workflow": [
            "image-edit"
          ]
        }
      },
      {
        "name": "image_flux3_klein_t2i",
        "title": "FLUX 3 Image Klein: Text to Image",
        "task": "Text to Image",
        "model": "FLUX 3 Image",
        "freshness": "new",
        "usage": 1394,
        "recommend": "top",
        "description": "Generate images from text prompts using FLUX 3 Image Klein, producing a single saved image with no file inputs required. Ideal for concept art, marketing visuals, and rapid prototyping of scene ideas from written descriptions.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.41.0",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "image_flux3_klein_distilled_t2i",
        "title": "FLUX 3 Image Klein Distilled: Text to Image",
        "task": "Text to Image",
        "model": "FLUX 3 Image",
        "freshness": "new",
        "usage": 211,
        "recommend": "medium",
        "description": "Generate images from text prompts using the distilled FLUX 3 Image Klein model, which runs efficiently in 15 sampling steps with no classifier-free guidance. Just describe what you want and receive one finished image. Ideal for rapid concept art, marketing visuals, and everyday creative ideation where speed matters.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.41.0",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "api_google_nano_banana2_image_edit",
        "title": "Nano Banana 2: Image Edit",
        "task": "Image Edit",
        "model": "Google",
        "freshness": "established",
        "usage": 2638,
        "recommend": "highly_recommended",
        "description": "Edit an image using natural language instructions with Google’s Nano Banana 2 model. Provide an input image and a text prompt describing the desired modification; the workflow generates an edited image that preserves subject consistency and high visual fidelity. This workflow executes via a third-party API.",
        "io": {
          "inputs": [
            "image: Input image to edit",
            "text: Text prompt describing the desired modification"
          ],
          "outputs": [
            "image: Edited image with consistent subject and high fidelity"
          ]
        },
        "minComfyUIVersion": "0.20.3",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ],
          "model_options": {
            "GeminiNanoBanana2V2": [
              "Gemini Nano Banana 2.1",
              "Nano Banana 2 (Gemini 3.1 Flash Image)",
              "Nano Banana 2 Lite"
            ]
          }
        }
      },
      {
        "name": "api_krea3_image_edit",
        "title": "Krea 3: Image Edit",
        "task": "Image Edit",
        "model": "Krea 3",
        "freshness": "new",
        "usage": 955,
        "recommend": "high",
        "description": "Perform targeted regional edits on an image with Krea 3, changing only the areas you specify while the rest of the picture stays untouched. One image goes in and one edited image comes out. Ideal for e-commerce product restaging, fashion and portrait relighting, and architectural sketch-to-render work.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.41.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ],
          "model_options": {
            "Krea3ImageNode": [
              "Krea 3"
            ]
          }
        }
      },
      {
        "name": "api_krea3_t2i",
        "title": "Krea 3: Text to Image",
        "task": "Text to Image",
        "model": "Krea 3",
        "freshness": "new",
        "usage": 359,
        "recommend": "medium",
        "description": "Generate images from text prompts using Krea 3, producing a single finished image per run with no input files required. Ideal for e-commerce product staging, fashion and portrait work, and architectural concept rendering.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.41.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-image"
          ],
          "model_options": {
            "Krea3ImageNode": [
              "Krea 3"
            ]
          }
        }
      },
      {
        "name": "api_bfl_flux3_image_edit",
        "title": "FLUX 3 Image: Image Edit",
        "task": "Image Edit",
        "model": "FLUX 3 Image",
        "freshness": "recent",
        "usage": 1317,
        "recommend": "top",
        "description": "Edit or combine up to ten reference images from a text instruction with FLUX 3 Image, referring to them in the prompt as image 1, image 2, and so on. One image is output at any of sixteen aspect ratios and up to 4K, with web and image search grounding on by default. Ideal for character-consistent edits, product and object placement, and multi-reference scene compositing.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.38.2",
        "capabilities": {
          "workflow": [
            "image-edit",
            "api"
          ]
        }
      },
      {
        "name": "api_bfl_flux3_t2i",
        "title": "FLUX 3 Image: Text to Image",
        "task": "Text to Image",
        "model": "FLUX 3 Image",
        "freshness": "recent",
        "usage": 346,
        "recommend": "medium",
        "description": "Generate an image from a text prompt with FLUX 3 Image, which interprets and expands the prompt before generating it, at any of sixteen aspect ratios and up to 4K. Web and image search grounding is on by default and no input files are required. One image is output. Ideal for typography-led posters, layout-driven design work, and fast exploration of visual concepts.\n",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.38.2",
        "capabilities": {
          "workflow": [
            "text-to-image",
            "api"
          ]
        }
      },
      {
        "name": "api_nano_banana_2_1_image_edit",
        "title": "Nano Banana 2.1: Image Edit",
        "task": "Image Edit",
        "model": "Nano Banana",
        "freshness": "new",
        "usage": 2271,
        "recommend": "top",
        "description": "Edit images with natural language and mask-based local edits powered by Nano Banana 2.1, taking one input image and returning one edited result alongside a side-by-side comparison view. Ideal for brand marketing posters, e-commerce product shots, and multi-round retouching where subjects and text must stay consistent.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.39.1",
        "capabilities": {
          "workflow": [
            "image-edit",
            "api"
          ],
          "model_options": {
            "GeminiNanoBanana2V2": [
              "Gemini Nano Banana 2.1",
              "Nano Banana 2 (Gemini 3.1 Flash Image)",
              "Nano Banana 2 Lite"
            ]
          }
        }
      },
      {
        "name": "api_nano_banana_pro",
        "title": "Nano Banana Pro",
        "task": "Image Edit",
        "model": "Gemini3 Pro Image Preview",
        "freshness": "established",
        "usage": 2709,
        "recommend": "highly_recommended",
        "description": "This workflow performs image editing using Google’s latest image model, supporting brand design, style references, and multiple angles. Users provide one or two reference images to generate or edit a studio-quality 4K output with enhanced text and character consistency. Execution requires a third-party API call.",
        "io": {
          "inputs": [
            "image×2: Reference images for generation or editing"
          ],
          "outputs": [
            "image: Generated or edited high-resolution image"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ],
          "model_options": {
            "GeminiImage2Node": [
              "gemini-3-pro-image-preview",
              "Nano Banana 2 (Gemini 3.1 Flash Image)"
            ]
          }
        }
      },
      {
        "name": "api_bfl_flux3_image_bboxes",
        "title": "FLUX 3 Image: Bounding Box Edit",
        "task": "Image Edit",
        "model": "FLUX 3 Image",
        "freshness": "recent",
        "usage": 160,
        "recommend": "low",
        "description": "Place objects and text where you want them with FLUX 3 Image by feeding boxes from the Create Bounding Boxes node into the prompt, so each box carries its own description and copy into the render. Boxes are relative to the canvas, so set the output to the same aspect ratio. One image is output, up to 4K. Ideal for poster and layout composition, placing headlines and captions, and controlling where subjects sit in the frame.\n",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.38.2",
        "capabilities": {
          "workflow": [
            "image-edit",
            "api"
          ]
        }
      },
      {
        "name": "api_ideogram_v4_5_precise_image_edit",
        "title": "Ideogram 4.5 Precise Edit",
        "task": "Image Edit",
        "model": "Ideogram",
        "freshness": "recent",
        "usage": 143,
        "recommend": "low",
        "description": "Generate a precisely edited image from a text prompt and up to five input images using Ideogram 4.5, where image 1 is the image to edit and images 2-5 are optional references. Only what the prompt asks for changes, so untouched pixels stay identical and the output keeps the size of image 1, with a choice of quality tiers. Ideal for product recolours and material swaps, replacing one element of a photo without disturbing its lighting, background, or framing, and touch-ups that must leave the rest of the image alone.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.43.2",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ]
        }
      },
      {
        "name": "api_ideogram_v4_5_image_edit",
        "title": "Ideogram 4.5: Image Edit",
        "task": "Image Edit",
        "model": "Ideogram",
        "freshness": "recent",
        "usage": 153,
        "recommend": "low",
        "description": "Generate edited or combined images from a text prompt and up to five input images using Ideogram 4.5, where image 1 is the image to edit and images 2-5 are references for design, material, or colour. Canvas size, quality level, and seed are adjustable, and one image is output. Ideal for product and outfit swaps, applying a reference design or palette to an existing photo, and recomposing a shot into a different aspect ratio.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.38.1",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ]
        }
      },
      {
        "name": "api_ideogram_v4_5_t2i",
        "title": "Ideogram 4.5: Text to Image",
        "task": "Text to Image",
        "model": "Ideogram",
        "freshness": "recent",
        "usage": 52,
        "recommend": "low",
        "description": "Generate images from text prompts using Ideogram 4.5, with controls for canvas size, quality level, automatic prompt rewriting, and seed. No file inputs are required and one image is output. Ideal for typography-led posters, layout-driven design work, marketing visuals, and fast exploration of visual concepts.\n",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.38.1",
        "capabilities": {
          "workflow": [
            "text-to-image",
            "api"
          ]
        }
      },
      {
        "name": "api_grok_image_edit",
        "title": "Grok: Image Edit",
        "task": "Image Edit",
        "model": "Grok",
        "freshness": "established",
        "usage": 2250,
        "recommend": "top",
        "description": "Upload an image to edit with the Grok model. Generate a modified version with expressive, cinematic visuals",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.11.1",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ],
          "model_options": {
            "GrokImageEditNodeV2": [
              "grok-imagine-image-2.0",
              "grok-imagine-image-quality",
              "grok-imagine-image-pro",
              "grok-imagine-image"
            ]
          }
        }
      },
      {
        "name": "api_bytedance_seedream_5_0_pro_image_edit",
        "title": "Seedream 5.0 Pro: Image Edit",
        "task": "Image Edit",
        "model": "Seedream 5.0 Pro",
        "freshness": "current",
        "usage": 3207,
        "recommend": "highly_recommended",
        "description": "Edit images with precision using Seedream 5.0 Pro, targeting specific regions while preserving lighting, depth, and texture. The workflow takes 1 input image and produces 1 output image. Ideal for product photography edits, portrait consistency adjustments, and creating structured layouts with multilingual text.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.31.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ],
          "model_options": {
            "ByteDanceSeedreamNodeV3": [
              "seedream 5.0 pro",
              "seedream 5.0 flash",
              "seedream 5.0 lite",
              "seedream-4-5-251128",
              "seedream-4-0-250828"
            ]
          }
        }
      },
      {
        "name": "api_bytedance_seedream_5_0_pro_t2i",
        "title": "Seedream 5.0 Pro: Text to Image",
        "task": "Text to Image",
        "model": "Seedream 5.0 Pro",
        "freshness": "current",
        "usage": 2937,
        "recommend": "highly_recommended",
        "description": "Generate images from text prompts using Seedream 5.0 Pro, producing one high-quality output. Ideal for creative concept art, marketing visual mockups, and rapid ideation for design projects.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.31.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-image"
          ],
          "model_options": {
            "ByteDanceSeedreamNodeV3": [
              "seedream 5.0 pro",
              "seedream 5.0 flash",
              "seedream 5.0 lite",
              "seedream-4-5-251128",
              "seedream-4-0-250828"
            ]
          }
        }
      },
      {
        "name": "image_flux2",
        "title": "FLUX.2 Dev",
        "task": "Image Edit",
        "model": "FLUX",
        "freshness": "established",
        "usage": 904,
        "recommend": "high",
        "description": "Generate photorealistic images with multi-reference consistency and professional text rendering.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.4.0",
        "capabilities": {
          "workflow": [
            "image-edit",
            "text-to-image"
          ]
        }
      },
      {
        "name": "image_flux2_klein_image_edit_9b_base",
        "title": "FLUX.2 [Klein] 9B: Image Edit",
        "task": "Image Edit",
        "model": "FLUX",
        "freshness": "established",
        "usage": 727,
        "recommend": "high",
        "description": "BFL undistilled foundation model. Maximum flexibility and control. Great for fine-tuning.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.3.72",
        "capabilities": {
          "workflow": [
            "image-edit"
          ]
        }
      },
      {
        "name": "api_openai_gpt_image_25_flare_t2i",
        "title": "GPT Image 2.5 Flare: Text to Image",
        "task": "Text to Image",
        "model": "GPT Image 2.5",
        "freshness": "recent",
        "usage": 934,
        "recommend": "high",
        "description": "GPT Image 2.5 Flare is a fast image generation and editing template optimized for rapid prototyping and creative iteration. It delivers quick previews for concept exploration, mood boards, poster drafts, character sketches and scene mockups. Balances solid visual quality with low latency, ideal for high-volume trial-and-error creative workflows.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.35.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-image"
          ],
          "model_options": {
            "OpenAIGPTImageNodeV2": [
              "gpt-image-2.5-flare",
              "gpt-image-2.5-sunburst",
              "gpt-image-2",
              "gpt-image-1.5",
              "gpt-image-1"
            ]
          }
        }
      },
      {
        "name": "image_qwen_image_edit_2511_int8",
        "title": "Qwen Image Edit 2511 Int8: Image Edit",
        "task": "Image Edit",
        "model": "Qwen Image Edit 2511",
        "freshness": "current",
        "usage": 1097,
        "recommend": "top",
        "description": "Edit images with Qwen Image Edit 2511 Int8, using a quantized model for faster inference while maintaining high quality. Input one image and receive one edited output. Ideal for character-consistent portrait editing, multi-person group photo fusion, and industrial design modifications.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.23.0",
        "capabilities": {
          "workflow": [
            "image-edit",
            "int8"
          ]
        }
      },
      {
        "name": "api_openai_gpt_image_25_flare_image_edit",
        "title": "GPT Image 2.5 Flare: Image Edit",
        "task": "Image Edit",
        "model": "GPT Image 2.5",
        "freshness": "recent",
        "usage": 914,
        "recommend": "high",
        "description": "Template for image rework & partial modification. Keep scene, pose, lighting and composition unchanged, only edit outfit, hairstyle or local details. Fast preview and iterative style adjustment powered by GPT-Image-2.5-Flare.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.35.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ],
          "model_options": {
            "OpenAIGPTImageNodeV2": [
              "gpt-image-2.5-flare",
              "gpt-image-2.5-sunburst",
              "gpt-image-2",
              "gpt-image-1.5",
              "gpt-image-1"
            ]
          }
        }
      },
      {
        "name": "api_openai_gpt_image_25_sunburst_image_edit",
        "title": "GPT Image 2.5 Sunburst: Image Edit",
        "task": "Image Edit",
        "model": "GPT Image 2.5",
        "freshness": "recent",
        "usage": 2331,
        "recommend": "top",
        "description": "High-fidelity image editing template. Preserve original character pose, scene, lighting and composition. Only modify partial elements such as outfit, hairstyle, facial features or camera angle. Great for sequential frame generation with strong subject consistency.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.35.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ],
          "model_options": {
            "OpenAIGPTImageNodeV2": [
              "gpt-image-2.5-flare",
              "gpt-image-2.5-sunburst",
              "gpt-image-2",
              "gpt-image-1.5",
              "gpt-image-1"
            ]
          }
        }
      },
      {
        "name": "api_bytedance_seedream4",
        "title": "Seedream 4.0: Image Edit",
        "task": "Image Edit",
        "model": "ByteDance",
        "freshness": "established",
        "usage": 858,
        "recommend": "high",
        "description": "Multi-modal AI model for text-to-image and image editing. Generate 2K images in under 2 seconds with natural language control.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit",
            "text-to-image"
          ],
          "model_options": {
            "ByteDanceSeedreamNodeV3": [
              "seedream 5.0 pro",
              "seedream 5.0 flash",
              "seedream 5.0 lite",
              "seedream-4-5-251128",
              "seedream-4-0-250828"
            ]
          }
        }
      },
      {
        "name": "api_bytedance_seedream_5_0_lite_image_edit",
        "title": "Seedream 5.0 Lite: Image Edit",
        "task": "Image Edit",
        "model": "Seedream 5.0",
        "freshness": "established",
        "usage": 787,
        "recommend": "high",
        "description": "Upload an image and a text prompt to edit it. Generate a modified image where the details align with your instructions.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.31.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ],
          "model_options": {
            "ByteDanceSeedreamNodeV3": [
              "seedream 5.0 pro",
              "seedream 5.0 flash",
              "seedream 5.0 lite",
              "seedream-4-5-251128",
              "seedream-4-0-250828"
            ]
          }
        }
      },
      {
        "name": "image_krea2_turbo_t2i_int8",
        "title": "Krea-2 Int8: Text to Image",
        "task": "Text to Image",
        "model": "Krea 2",
        "freshness": "current",
        "usage": 667,
        "recommend": "high",
        "description": "Generate images with the Krea-2 Turbo distilled text-to-image model using the high-performance Int8 Convrot format, which delivers better quality than FP8 while generally running faster across most GPUs. This workflow accepts no file inputs and produces a single image output, with optional style LoRA support for creative customization. Ideal for rapid concept art iteration, batch asset generation for games or marketing, and exploring the speed-quality sweet spot of the Convrot Int8 quantization.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.24.1",
        "capabilities": {
          "workflow": [
            "text-to-image",
            "int8"
          ]
        }
      },
      {
        "name": "image_krea2_turbo_int8_image_style_reference",
        "title": "Krea-2 Int8: Image Style Reference",
        "task": "Style Reference",
        "model": "Krea 2",
        "freshness": "current",
        "usage": 353,
        "recommend": "medium",
        "description": "Generate images with the Krea-2 Turbo model while referencing the style of 1–2 uploaded images, using the high-performance Int8 Convrot format for faster, higher-quality inference. This workflow accepts up to two reference image inputs and produces a single stylized output, with optional LoRA support for further creative control. Ideal for matching a specific artistic style in concept art, producing consistent brand visuals, and exploring rapid style transfer for iterative design.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.24.1",
        "capabilities": {
          "workflow": [
            "style-reference",
            "int8"
          ]
        }
      },
      {
        "name": "api_bytedance_seedream_5_0_flash_image_edit",
        "title": "Seedream 5.0 Flash: Image Edit",
        "task": "Image Edit",
        "model": "Seedream 5.0",
        "freshness": "recent",
        "usage": 275,
        "recommend": "medium",
        "description": "Edit and generate images with Seedream 5.0 Flash, transforming a single input image into one refined output with cinematic lighting, accurate text rendering, and resolutions up to 2K.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.37.3",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ],
          "model_options": {
            "ByteDanceSeedreamNodeV3": [
              "seedream 5.0 pro",
              "seedream 5.0 flash",
              "seedream 5.0 lite",
              "seedream-4-5-251128",
              "seedream-4-0-250828"
            ]
          }
        }
      },
      {
        "name": "image_qwen_image_edit_2511",
        "title": "Qwen Image Edit 2511 - Material Replacement",
        "task": "Image Edit",
        "model": "Qwen-Image-Edit",
        "freshness": "established",
        "usage": 432,
        "recommend": "medium",
        "description": "Replace materials in objects (e.g., furniture) by combining reference images with Qwen-Image-Edit-2511.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.4.0",
        "capabilities": {
          "workflow": [
            "image-edit"
          ]
        }
      },
      {
        "name": "image_z_image_turbo_int8",
        "title": "Z-Image-Turbo Int8: Text to Image",
        "task": "Text to Image",
        "model": "Z-Image-Turbo",
        "freshness": "current",
        "usage": 90,
        "recommend": "low",
        "description": "Generate photorealistic images using the Z-Image-Turbo int8 model, a distilled 6B-parameter variant optimized for sub-second inference through efficient 8-step sampling. The workflow takes a text prompt as input and outputs a single high-quality image. Ideal for rapid creative iteration, realistic concept art, and production-ready visual prototyping.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.11.0",
        "capabilities": {
          "workflow": [
            "text-to-image",
            "int8"
          ]
        }
      },
      {
        "name": "image_firered_image_edit1_1",
        "title": "FireRed Image Edit 1.1: Image Edit",
        "task": "Image Edit",
        "model": "FireRed Image Edit 1.1",
        "freshness": "established",
        "usage": 762,
        "recommend": "high",
        "description": "Upload an image and a text prompt to edit it. Generate a modified image with enhanced identity consistency, multi-element fusion, or professional photo restoration.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "image-edit"
          ]
        }
      },
      {
        "name": "api_grok_text_to_image",
        "title": "Grok: Text to Image",
        "task": "Text to Image",
        "model": "Grok",
        "freshness": "established",
        "usage": 273,
        "recommend": "medium",
        "description": "Use the Grok API to generate high-quality images from text prompts. This workflow takes a text description as input and produces a generated image. Grok enables fast inference and versatile image creation directly through the API.",
        "io": {
          "inputs": [
            "text: Text prompt describing the desired image"
          ],
          "outputs": [
            "image: Generated image from text prompt"
          ]
        },
        "minComfyUIVersion": "0.11.1",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-image"
          ],
          "model_options": {
            "GrokImageNode": [
              "grok-imagine-image-2.0",
              "grok-imagine-image-quality",
              "grok-imagine-image-pro",
              "grok-imagine-image"
            ]
          }
        }
      },
      {
        "name": "api_grok_imagine_image_2_image_edit",
        "title": "Grok Imagine Image 2.0: Image Edit",
        "task": "Image Edit",
        "model": "Grok Imagine Image 2.0",
        "freshness": "recent",
        "usage": 575,
        "recommend": "high",
        "description": "Grok Imagine Image 2.0 performs precise image editing and text rendering using xAI's Aurora engine, letting you modify specific elements, swap backgrounds, or restyle scenes through natural language with crisp short-text output. It takes up to 3 reference images and generates at 1K or 2K resolution across 14 aspect ratios, with strong character consistency across scenes. Ideal for poster and cover design, product and brand visuals, or multi-scene character workflows.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.32.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ],
          "model_options": {
            "GrokImageEditNodeV2": [
              "grok-imagine-image-2.0",
              "grok-imagine-image-quality",
              "grok-imagine-image-pro",
              "grok-imagine-image"
            ]
          }
        }
      },
      {
        "name": "api_grok_imagine_image_2_t2i",
        "title": "Grok Imagine Image 2.0: Text to Image",
        "task": "Text to Image",
        "model": "Grok Imagine Image 2.0",
        "freshness": "recent",
        "usage": 339,
        "recommend": "medium",
        "description": "Generate images from text prompts using Grok Imagine Image 2.0, an Aurora-engine model that excels at crisp short-text rendering and 1K or 2K outputs across 14 aspect ratios. Enter one text prompt and receive one generated image, with low or medium quality options to control cost. Ideal for poster and cover design, product and brand visuals, and character-driven scenes where consistency matters.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.32.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-image"
          ],
          "model_options": {
            "GrokImageNode": [
              "grok-imagine-image-2.0",
              "grok-imagine-image-quality",
              "grok-imagine-image-pro",
              "grok-imagine-image"
            ]
          }
        }
      },
      {
        "name": "image_flux2_text_to_image",
        "title": "FLUX.2 Dev Text to Image",
        "task": "Text to Image",
        "model": "FLUX",
        "freshness": "established",
        "usage": 403,
        "recommend": "medium",
        "description": "Text-to-image with enhanced lighting, materials, and realistic details.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.4.0",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "image_flux2_klein_image_edit_4b_base",
        "title": "FLUX.2 [Klein] 4B: Image Edit",
        "task": "Image Edit",
        "model": "FLUX",
        "freshness": "established",
        "usage": 282,
        "recommend": "medium",
        "description": "A smaller foundation model with exceptional quality-to-size ratio. Ideal for local deployment, fine-tuning on limited hardware, and efficient generation and editing workflows.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.3.72",
        "capabilities": {
          "workflow": [
            "image-edit"
          ]
        }
      },
      {
        "name": "image_flux2_text_to_image_9b",
        "title": "FLUX.2 [Klein] 9B: Text to Image",
        "task": "Text to Image",
        "model": "FLUX",
        "freshness": "established",
        "usage": 244,
        "recommend": "medium",
        "description": "BFL distilled model. Outstanding quality at sub-second speed. Great for real-time generation while retaining quality. Marketing launch will focus on this model.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.3.72",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "image_flux2_klein_image_edit_4b_distilled",
        "title": "FLUX.2 [Klein] 4B Distilled: Image Edit",
        "task": "Image Edit",
        "model": "FLUX",
        "freshness": "established",
        "usage": 300,
        "recommend": "medium",
        "description": "The fastest variant in the Klein family. Built for interactive applications, real-time previews, and latency-critical production use cases.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.3.72",
        "capabilities": {
          "workflow": [
            "image-edit"
          ]
        }
      },
      {
        "name": "flux_fill_inpaint_example",
        "title": "Flux.1 Inpaint",
        "task": "Image Inpainting",
        "model": "FLUX",
        "freshness": "established",
        "usage": 564,
        "recommend": "high",
        "description": "Fill missing parts of images using Flux.1 Fill Inpainting.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.42",
        "capabilities": {
          "workflow": [
            "inpainting"
          ]
        }
      },
      {
        "name": "flux_kontext_dev_basic",
        "title": "Flux Kontext Dev Image Edit",
        "task": "Image Edit",
        "model": "FLUX.1 Kontext",
        "freshness": "established",
        "usage": 262,
        "recommend": "medium",
        "description": "Smart image editing that keeps characters consistent, edits specific parts without affecting others, and preserves original styles.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.42",
        "capabilities": {
          "workflow": [
            "image-edit"
          ]
        }
      },
      {
        "name": "image_flux2_klein_text_to_image",
        "title": "FLUX.2 [Klein] 4B: Text to Image",
        "task": "Text to Image",
        "model": "FLUX",
        "freshness": "established",
        "usage": 108,
        "recommend": "low",
        "description": "BFL distilled model. Outstanding quality at sub-second speed. Great for real-time generation while retaining quality. Marketing launch will focus on this model.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.3.72",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "image_anima_base_v1",
        "title": "Anima Base v1: Text to Image",
        "task": "Text to Image",
        "model": "Anima",
        "freshness": "current",
        "usage": 734,
        "recommend": "high",
        "description": "Input a text prompt describing an anime or artistic illustration. Generate a non-photorealistic image focused on anime concepts, characters, or styles.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.16.0",
        "capabilities": {
          "workflow": [
            "anime",
            "text-to-image"
          ]
        }
      },
      {
        "name": "image_anima_preview",
        "title": "Anima Anime Text-to-Image Generation",
        "task": "Text to Image",
        "model": "Anima",
        "freshness": "established",
        "usage": 270,
        "recommend": "medium",
        "description": "Input a text prompt to generate an anime-style image using the Anima model. Configure settings like steps and CFG scale to control the output. The workflow produces a non-photorealistic illustration based on your description.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.11.0",
        "capabilities": {
          "workflow": [
            "anime",
            "text-to-image"
          ]
        }
      },
      {
        "name": "api_google_nano_banana2_text_to_image",
        "title": "Nano Banana 2: Text to Image",
        "task": "Text to Image",
        "model": "Nano Banana 2",
        "freshness": "current",
        "usage": 131,
        "recommend": "low",
        "description": "Generate images from text prompts using Google's Gemini 3.1 Flash Image model (Nano Banana 2), optimized for fast generation and up to 4K resolution. This API workflow takes a text prompt and outputs a single image, ideal for rapid prototyping, concept art, and creative exploration.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.20.3",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-image"
          ],
          "model_options": {
            "GeminiNanoBanana2V2": [
              "Gemini Nano Banana 2.1",
              "Nano Banana 2 (Gemini 3.1 Flash Image)",
              "Nano Banana 2 Lite"
            ]
          }
        }
      },
      {
        "name": "api_nano_banana_2_lite_image_edit",
        "title": "Nano Banana 2 Lite: Image Edit",
        "task": "Image Edit",
        "model": "Nano Banana 2",
        "freshness": "current",
        "usage": 138,
        "recommend": "low",
        "description": "Edit an image using Nano Banana 2 Lite, Google’s fastest Gemini flash model optimized for rapid generation up to 4K. It accepts a text prompt and outputs a side-by-side comparison of the original and edited result. This workflow supports API execution for programmatic use.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.27.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ],
          "model_options": {
            "GeminiNanoBanana2V2": [
              "Gemini Nano Banana 2.1",
              "Nano Banana 2 (Gemini 3.1 Flash Image)",
              "Nano Banana 2 Lite"
            ]
          }
        }
      },
      {
        "name": "api_openai_gpt_image_2_image_edit",
        "title": "GPT Image 2: Image Edit",
        "task": "Image Edit",
        "model": "GPT Image 2",
        "freshness": "established",
        "usage": 192,
        "recommend": "low",
        "description": "Use GPT Image 2 to edit an image via API by providing a text prompt. The model applies changes like adding objects or altering style while maintaining high prompt adherence. Upload a reference image and describe the desired edit to produce photorealistic or commercial-ready outputs.",
        "io": {
          "inputs": [
            "image: Reference image for editing",
            "text: Text prompt describing the desired edit or design"
          ],
          "outputs": [
            "image: Generated commercial-quality image with clean text rendering"
          ]
        },
        "minComfyUIVersion": "0.21.1",
        "capabilities": {
          "workflow": [
            "image-edit",
            "api"
          ],
          "model_options": {
            "OpenAIGPTImageNodeV2": [
              "gpt-image-2.5-flare",
              "gpt-image-2.5-sunburst",
              "gpt-image-2",
              "gpt-image-1.5",
              "gpt-image-1"
            ]
          }
        }
      },
      {
        "name": "api_openai_gpt_image_2_t2i",
        "title": "GPT Image 2: Text to Image",
        "task": "Text to Image",
        "model": "GPT Image 2",
        "freshness": "established",
        "usage": 253,
        "recommend": "medium",
        "description": "Generate a detailed image from a text prompt using OpenAI's GPT Image 2, known for superior prompt adherence and strong character generation. This workflow accepts only a text prompt and outputs a high-quality image, leveraging the model's text-to-image capability via API.",
        "io": {
          "inputs": [
            "text: Text prompt describing the desired image or design"
          ],
          "outputs": [
            "image: Generated photorealistic commercial image"
          ]
        },
        "minComfyUIVersion": "0.21.1",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-image"
          ],
          "model_options": {
            "OpenAIGPTImageNodeV2": [
              "gpt-image-2.5-flare",
              "gpt-image-2.5-sunburst",
              "gpt-image-2",
              "gpt-image-1.5",
              "gpt-image-1"
            ]
          }
        }
      },
      {
        "name": "api_openai_gpt_image_25_sunburst_t2i",
        "title": "GPT Image 2.5 Sunburst: Text to Image",
        "task": "Text to Image",
        "model": "GPT Image 2.5",
        "freshness": "recent",
        "usage": 629,
        "recommend": "high",
        "description": "GPT Image 2.5 Sunburst is a high-fidelity image editing and generation template for creating polished, detail-rich visuals. It preserves subject identity, complex materials, lighting consistency and scene structure across iterations. Suitable for concept refinement, product visuals, character scenes, cinematic backgrounds and finished artwork that requires precision, realism and visual depth.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.35.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-image"
          ],
          "model_options": {
            "OpenAIGPTImageNodeV2": [
              "gpt-image-2.5-flare",
              "gpt-image-2.5-sunburst",
              "gpt-image-2",
              "gpt-image-1.5",
              "gpt-image-1"
            ]
          }
        }
      },
      {
        "name": "image_sdxl_simple",
        "title": "SDXL1.0: Text to Image (Simple)",
        "task": "Text to Image",
        "model": "SDXL",
        "freshness": "established",
        "usage": 456,
        "recommend": "medium",
        "description": "Generate high-quality images from text prompts using the SDXL 1.0 model, offering a straightforward text-to-image generation workflow. Ideal for quick concept visualization, creative brainstorming, and rapid prototyping of visual ideas.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.0.1",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "api_tencent_hy_image_3_5_preview_image_edit",
        "title": "HY Image 3.5 Preview: Image Edit",
        "task": "Image Edit",
        "model": "HY Image 3.5",
        "freshness": "recent",
        "usage": 22,
        "recommend": "not_recommended",
        "description": "Edit images with the HY Image 3.5 Preview model, which deeply understands user intent through optimized text-and-image organization to deliver high-quality results. Takes 1 image input and produces 1 image output. Ideal for portrait editing, commercial posters, and film/TV marketing materials.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.37.1",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ]
        }
      },
      {
        "name": "api_tencent_hy_image_3_5_preview_t2i",
        "title": "HY Image 3.5 Preview: Text to Image",
        "task": "Text to Image",
        "model": "HY Image 3.5",
        "freshness": "recent",
        "usage": 15,
        "recommend": "not_recommended",
        "description": "Tencent Hunyuan's preview image model, strengthening three areas over previous versions: multilingual text rendering and typesetting for advertising, infographics and UI design; photorealism and stylistic range covering cinematic, comic, commercial photography and illustration; and editing consistency that preserves key traits of people and products across background, outfit and style changes. Outputs up to 4K resolution.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.37.1",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-image"
          ]
        }
      },
      {
        "name": "image_qwen_Image_2512",
        "title": "Qwen Image 2512",
        "task": "Text to Image",
        "model": "Qwen",
        "freshness": "established",
        "usage": 273,
        "recommend": "medium",
        "description": "Text-to-image model with enhanced human realism, finer natural details for landscapes and animal fur, and improved text rendering with accurate layout and multimodal composition.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.4.0",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "image_qwen_image_instantx_inpainting_controlnet",
        "title": "Qwen-Image InstantX Inpainting ControlNet",
        "task": "Image Inpainting",
        "model": "Qwen-Image",
        "freshness": "established",
        "usage": 185,
        "recommend": "low",
        "description": "Professional inpainting and image editing with Qwen-Image InstantX ControlNet. Supports object replacement, text modification, background changes, and outpainting.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.57",
        "capabilities": {
          "workflow": [
            "controlnet",
            "inpainting"
          ]
        }
      },
      {
        "name": "api_qwen3_image_edit",
        "title": "Qwen Image 3.0 Pro: Image Edit",
        "task": "Image Edit",
        "model": "Qwen Image 3.0 Pro",
        "freshness": "current",
        "usage": 110,
        "recommend": "low",
        "description": "Qwen Image 3.0 Pro performs advanced image editing with rich multi-image input, generating complex layouts like newspapers, storyboards, menus, and exam papers in a single pass while preserving fine details down to 10px text, micro-expressions, and individual hairs. Supports native rendering across 12 languages with realistic simulation of web pages, games, and live streams. Ideal for professional content production, interface mockups, and high-fidelity commercial imagery requiring authentic detail.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.32.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ]
        }
      },
      {
        "name": "api_qwen3_t2i",
        "title": "Qwen Image 3.0 Pro: Text to Image",
        "task": "Text to Image",
        "model": "Qwen Image 3.0 Pro",
        "freshness": "current",
        "usage": 176,
        "recommend": "low",
        "description": "Generate detailed, production-ready images from text prompts using the Qwen Image 3.0 Pro model, supporting up to 4.5k tokens for complex layouts like posters, infographics, and storyboards. It renders crisp 10px text, realistic textures, and native support for 12 languages and over 100 art styles, with optional 1-3 reference images for fusion editing. Ideal for marketing materials, educational diagrams, and UI design mockups.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.32.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-image"
          ]
        }
      },
      {
        "name": "api_bytedance_seedream_5_0_flash_t2i",
        "title": "Seedream 5.0 Flash: Text to Image",
        "task": "Text to Image",
        "model": "Seedream 5.0",
        "freshness": "recent",
        "usage": 139,
        "recommend": "low",
        "description": "Generate high-quality images from text prompts using Seedream 5.0 Flash, producing detailed 1K, 1.5K, or 2K visuals with cinematic lighting and accurate text rendering, saved as a single image output.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.37.3",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-image"
          ],
          "model_options": {
            "ByteDanceSeedreamNodeV3": [
              "seedream 5.0 pro",
              "seedream 5.0 flash",
              "seedream 5.0 lite",
              "seedream-4-5-251128",
              "seedream-4-0-250828"
            ]
          }
        }
      },
      {
        "name": "flux1_dev_uso_reference_image_gen",
        "title": "Flux.1 Dev USO Reference Image Generation",
        "task": "Style Reference",
        "model": "FLUX",
        "freshness": "established",
        "usage": 214,
        "recommend": "medium",
        "description": "Use reference images to control both style and subject - keep your character's face while changing artistic style, or apply artistic styles to new scenes",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.57",
        "capabilities": {
          "workflow": [
            "style-reference"
          ]
        }
      },
      {
        "name": "flux1_krea_dev",
        "title": "Flux.1 Krea Dev",
        "task": "Text to Image",
        "model": "FLUX",
        "freshness": "established",
        "usage": 105,
        "recommend": "low",
        "description": "A fine-tuned FLUX model pushing photorealism to the max",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.42",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "api_flux2",
        "title": "FLUX.2 Pro",
        "task": "Image Edit",
        "model": "FLUX.2",
        "freshness": "established",
        "usage": 98,
        "recommend": "low",
        "description": "Generate up to 4MP photorealistic images with multi-reference consistency and professional text rendering.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "image-edit",
            "api",
            "text-to-image"
          ],
          "model_options": {
            "Flux2ImageNode": [
              "Flux.2 [pro]",
              "Flux.2 [max]"
            ]
          }
        }
      },
      {
        "name": "flux_dev_full_text_to_image",
        "title": "Flux.1 Dev: Text to Image",
        "task": "Text to Image",
        "model": "FLUX",
        "freshness": "established",
        "usage": 178,
        "recommend": "low",
        "description": "Generate high-quality images with Flux Dev full version. Requires larger VRAM and multiple model files, but provides the best prompt following capability and image quality.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.3.42",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "flux_canny_model_example",
        "title": "Flux.1 Canny Model",
        "task": "Image Generation",
        "model": "FLUX.1",
        "freshness": "established",
        "usage": 151,
        "recommend": "low",
        "description": "Generate images guided by edge detection using Flux.1 Canny.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.0.4",
        "capabilities": {
          "workflow": [
            "controlnet"
          ]
        }
      },
      {
        "name": "flux_dev_checkpoint_example",
        "title": "Flux.1 Dev fp8: Text to Image",
        "task": "Text to Image",
        "model": "FLUX.1",
        "freshness": "established",
        "usage": 152,
        "recommend": "low",
        "description": "Generate images using Flux.1 Dev fp8 quantized version. Suitable for devices with limited VRAM, requires only one model file, but image quality is slightly lower than the full version.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.42",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "api_kling_vto",
        "title": "Kling: Virtual Try On",
        "task": "Virtual Try-On",
        "model": "Kling",
        "freshness": "new",
        "usage": 23,
        "recommend": "medium",
        "description": "Dress a virtual model in any garment by combining a clothing image with a person photo, generating a realistic try-on result alongside a side-by-side comparison view. It accepts flat-lay, mannequin, or on-model clothing shots while preserving the model's face and pose. Ideal for fashion e-commerce previews, styling mockups, and virtual fitting experiences.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.39.2",
        "capabilities": {
          "workflow": [
            "virtual-try-on"
          ]
        }
      },
      {
        "name": "image_flux2_fp8",
        "title": "Product Mockup(FLUX.2 Dev FP8)",
        "task": "Text to Image",
        "model": "FLUX.2 Dev",
        "freshness": "established",
        "usage": 100,
        "recommend": "low",
        "description": "Create product mockups by applying design patterns to packaging, mugs, and other products using multi-reference consistency.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.4.0",
        "capabilities": {
          "workflow": [
            "text-to-image",
            "image-edit"
          ]
        }
      },
      {
        "name": "flux_fill_outpaint_example",
        "title": "Flux.1 Outpaint",
        "task": "Image Generation",
        "model": "FLUX.1",
        "freshness": "established",
        "usage": 144,
        "recommend": "low",
        "description": "Extend images beyond boundaries using Flux.1 outpainting.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.0.4",
        "capabilities": {
          "workflow": [
            "outpainting"
          ]
        }
      },
      {
        "name": "image_z_image",
        "title": "Z-Image: Text to Image",
        "task": "Text to Image",
        "model": "Z-Image",
        "freshness": "established",
        "usage": 318,
        "recommend": "medium",
        "description": "Foundation for creative freedom. Diverse aesthetics with exceptional photorealistic quality; ideal for fine-tuning; responsive to negative prompts; high generation diversity.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.23.0",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "api_recraft_v4_1_t2i",
        "title": "Recraft 4.1: Text to Image",
        "task": "Text to Image",
        "model": "Recraft",
        "freshness": "current",
        "usage": 142,
        "recommend": "low",
        "description": "Generate photorealistic and illustrative images from short text prompts using Recraft V4.1 via API. This text-to-image workflow produces one high-quality image per generation, with Standard and Pro modes for sharp details and consistent style. Ideal for concept art, logo and typography design, and lifestyle product mockups.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.29.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-image"
          ],
          "model_options": {
            "RecraftV4TextToImageNode": [
              "recraftv4_1",
              "recraftv4_1_flash",
              "recraftv4_1_utility",
              "recraftv4_1_pro",
              "recraftv4_1_utility_pro",
              "recraftv4",
              "recraftv4_pro",
              "recraftv4_styles",
              "recraftv4_styles_pro"
            ]
          }
        }
      },
      {
        "name": "api_recraft_v4_1_flash_t2i",
        "title": "Recraft 4.1 flash: Text to Image",
        "task": "Text to Image",
        "model": "Recraft",
        "freshness": "recent",
        "usage": 124,
        "recommend": "low",
        "description": "Generate images from text prompts using Recraft V4.1 Flash, a fast, low-cost model that produces one image output with no file inputs required. Ideal for high-volume generation, rapid iteration, photography and portraits, atmospheric scenes, and quick logo concept exploration.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.37.2",
        "capabilities": {
          "workflow": [
            "api"
          ],
          "model_options": {
            "RecraftV4TextToImageNode": [
              "recraftv4_1",
              "recraftv4_1_flash",
              "recraftv4_1_utility",
              "recraftv4_1_pro",
              "recraftv4_1_utility_pro",
              "recraftv4",
              "recraftv4_pro",
              "recraftv4_styles",
              "recraftv4_styles_pro"
            ]
          }
        }
      },
      {
        "name": "image_ideogram4_t2i",
        "title": "Ideogram v4: Text to Image",
        "task": "Text to Image",
        "model": "Ideogram",
        "freshness": "current",
        "usage": 26,
        "recommend": "not_recommended",
        "description": "Input a text prompt or structured JSON description. Generate an image with precise layout, color, and style control using Ideogram 4.0.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.24.0",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "image_ideogram4_t2i_int8",
        "title": "Ideogram v4 Int8: Text to Image",
        "task": "Text to Image",
        "model": "Ideogram",
        "freshness": "current",
        "usage": 24,
        "recommend": "not_recommended",
        "description": "Generate images from text prompts using the Ideogram v4 model, a state-of-the-art open-weight text-to-image system with exceptional text rendering and layout control. No file inputs are required, producing 1 image output. Ideal for creating marketing visuals, generating design mockups with precise typography, and producing high-resolution artwork with controlled compositions.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.24.0",
        "capabilities": {
          "workflow": [
            "text-to-image",
            "int8"
          ]
        }
      },
      {
        "name": "image_netayume_lumina_t2i",
        "title": "NetaYume Lumina Text to Image",
        "task": "Text to Image",
        "model": "OmniGen",
        "freshness": "established",
        "usage": 80,
        "recommend": "low",
        "description": "High-quality anime-style image generation with enhanced character understanding and detailed textures. Fine-tuned from Neta Lumina on Danbooru dataset.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.3.42",
        "capabilities": {
          "workflow": [
            "anime",
            "text-to-image"
          ]
        }
      },
      {
        "name": "image_z_image_turbo_fun_union_controlnet",
        "title": "Z-Image-Turbo Fun Union ControlNet",
        "task": "Image Generation",
        "model": "Z-Image-Turbo",
        "freshness": "established",
        "usage": 356,
        "recommend": "medium",
        "description": "Multi-control ControlNet supporting Canny, HED, Depth, Pose, and MLSD for Z-Image-Turbo.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.57",
        "capabilities": {
          "workflow": [
            "controlnet"
          ]
        }
      },
      {
        "name": "image_mage_flow_t2i_int8",
        "title": "Mage-Flow: Text to Image",
        "task": "Text to Image",
        "model": "Mage-Flow",
        "freshness": "current",
        "usage": 14,
        "recommend": "not_recommended",
        "description": "Generate high-resolution images from text prompts using Mage-Flow, a compact 4B-parameter diffusion model that outputs a single image at native resolutions from 512 to 2048 pixels per side. Ideal for rapid concept visualization, creative asset generation, and research in controlled image synthesis under limited compute budgets.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.29.0",
        "capabilities": {
          "workflow": [
            "text-to-image",
            "int8"
          ]
        }
      },
      {
        "name": "image_mage_flow_edit_turbo_int8",
        "title": "Mage-Flow-Edit Turbo: Image Edit",
        "task": "Image Edit",
        "model": "Mage-Flow",
        "freshness": "current",
        "usage": 30,
        "recommend": "not_recommended",
        "description": "Edit images using a reference photo and text instructions with the Mage-Flow-Edit Turbo model, a distilled 4B-parameter model that produces results in just 4 steps. It takes 1-3 reference images and a text prompt as input, then outputs an edited image matching your instructions. Ideal for quick style transfers, background replacements, and interactive photo editing sessions.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.29.0",
        "capabilities": {
          "workflow": [
            "image-edit",
            "int8"
          ]
        }
      },
      {
        "name": "image_mage_flow_turbo_t2i_int8",
        "title": "Mage-Flow Turbo: Text to Image",
        "task": "Text to Image",
        "model": "Mage-Flow",
        "freshness": "current",
        "usage": 13,
        "recommend": "not_recommended",
        "description": "Generate high-quality images from text prompts using the compact 4B-parameter Mage-Flow Turbo model, producing a single image output without any file inputs. This workflow supports native-resolution generation from 512 to 2048 pixels per side at any aspect ratio, including extreme 4:1 dimensions. Ideal for rapid concept visualization, creative asset generation, and low-latency image prototyping on modest hardware.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.29.0",
        "capabilities": {
          "workflow": [
            "text-to-image",
            "int8",
            "turbo"
          ]
        }
      },
      {
        "name": "image_qwen_image_edit",
        "title": "Qwen Image Edit",
        "task": "Image Edit",
        "model": "Qwen-Image-Edit",
        "freshness": "established",
        "usage": 68,
        "recommend": "low",
        "description": "Edit images with precise bilingual text editing and dual semantic/appearance editing capabilities using Qwen-Image-Edit's 20B MMDiT model.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.4.0",
        "capabilities": {
          "workflow": [
            "image-edit"
          ]
        }
      },
      {
        "name": "templates-image_to_real",
        "title": "Illustration to Realism",
        "task": "Style Transfer",
        "model": "Qwen-Image-Edit",
        "freshness": "established",
        "usage": 52,
        "recommend": "low",
        "description": "Input an illustration and generate a hyper realistic version using Qwen Image Edit 2509.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.60",
        "capabilities": {
          "workflow": [
            "style-transfer"
          ]
        }
      },
      {
        "name": "image_qwen_image_edit_2509_relight",
        "title": "Image Relight",
        "task": "Image Edit",
        "model": "Qwen-Image-Edit",
        "freshness": "established",
        "usage": 44,
        "recommend": "not_recommended",
        "description": "Relight images using Qwen-Image-Edit with LoRA support.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.3.60",
        "capabilities": {
          "workflow": [
            "image-edit",
            "relight"
          ]
        }
      },
      {
        "name": "api_krea2_style_reference",
        "title": "Krea 2: Style Reference",
        "task": "Style Reference",
        "model": "Krea 2",
        "freshness": "current",
        "usage": 18,
        "recommend": "not_recommended",
        "description": "Upload up to two style reference images and a text prompt to generate an image that matches the prompt while adopting the reference’s aesthetic. Krea 2 excels at artistic photorealism and realistic textures, making this workflow ideal for creative style transfer.",
        "io": {
          "inputs": [
            "image×2: Style reference images",
            "text: Text prompt describing the desired output"
          ],
          "outputs": [
            "image: Generated output image"
          ]
        },
        "minComfyUIVersion": "0.24.1",
        "capabilities": {
          "workflow": [
            "api",
            "style-reference"
          ]
        }
      },
      {
        "name": "api_krea2_t2i",
        "title": "Krea2: Text to Image",
        "task": "Text to Image",
        "model": "Krea 2",
        "freshness": "current",
        "usage": 37,
        "recommend": "not_recommended",
        "description": "Input a text prompt describing your desired scene. Krea 2 excels at artistic photorealism with realistic textures, generating four creative, aesthetic images via a third-party API. This workflow is ideal for quickly iterating on expressive illustrations or graphic design concepts.",
        "io": {
          "inputs": [
            "text: Text prompt describing the desired scene"
          ],
          "outputs": [
            "image×4: Generated images with creative variety"
          ]
        },
        "minComfyUIVersion": "0.24.1",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-image"
          ]
        }
      },
      {
        "name": "api_google_nano_banana_create_turnaround_sheet",
        "title": "Character Turnaround Sheet Generator",
        "task": "Image Edit",
        "model": "Nano Banana",
        "freshness": "recent",
        "usage": 118,
        "recommend": "low",
        "description": "Create orthographic multi-view character turnaround sheet from one reference image, keep consistent character features across front, side and back views, for sci-fi mech and armor character design.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.34.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ],
          "model_options": {
            "GeminiImage2Node": [
              "gemini-3-pro-image-preview",
              "Nano Banana 2 (Gemini 3.1 Flash Image)"
            ]
          }
        }
      },
      {
        "name": "api_google_gemini_image",
        "title": "Nano Banana",
        "task": "Image Edit",
        "model": "Gemini-2.5-Flash",
        "freshness": "established",
        "usage": 87,
        "recommend": "low",
        "description": "This workflow uses Google Gemini 2.5 Flash via API to perform image editing with consistency. It accepts a text prompt describing the edit and outputs an image. The model's fast generation and multimodal understanding enable dynamic visual content creation.",
        "io": {
          "inputs": [
            "image: Reference image to edit or use as generation basis",
            "text: Text prompt describing the desired edit"
          ],
          "outputs": [
            "image: Generated or edited image"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "image-edit",
            "api",
            "text-to-image"
          ],
          "model_options": {
            "GeminiImageNode": [
              "gemini-2.5-flash-image"
            ]
          }
        }
      },
      {
        "name": "image_boogu_image_0_1_edit",
        "title": "Boogu image 0.1 Edit",
        "task": "Image Edit",
        "model": "Boogu",
        "freshness": "current",
        "usage": 27,
        "recommend": "not_recommended",
        "description": "Edit images using Boogu's instruction-driven model, taking one input image and generating an edited output alongside a side-by-side comparison view.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.26.0",
        "capabilities": {
          "workflow": [
            "image-edit"
          ]
        }
      },
      {
        "name": "api_kling_o3_image",
        "title": "Kling O3: Image Generation",
        "task": "Image Edit",
        "model": "Kling O3",
        "freshness": "established",
        "usage": 61,
        "recommend": "low",
        "description": "Upload up to 10 reference images to guide the synthesis. Generate a new image using the Kling O3 model based on your provided visual inputs.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit",
            "text-to-image"
          ]
        }
      },
      {
        "name": "api_openrouter_mai_image_26_image_edit",
        "title": "MAI Image 2.6: Image Edit",
        "task": "Image Edit",
        "model": "Mai Image 2.6",
        "freshness": "recent",
        "usage": 13,
        "recommend": "not_recommended",
        "description": "Edit images with MAI Image 2.6, Microsoft's flagship model for image editing, combining two reference images into a single refined output. Ideal for multi-reference compositing, product visual restyling, and text-in-image design work.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.35.1",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ]
        }
      },
      {
        "name": "api_openrouter_mai_image_26_t2i",
        "title": "Mai Image 2.6: Text to Image",
        "task": "Text to Image",
        "model": "Mai Image 2.6",
        "freshness": "recent",
        "usage": 20,
        "recommend": "not_recommended",
        "description": "Generate high-quality images from text prompts using Microsoft's MAI Image 2.6, a top-tier model excelling in legible in-image text rendering, photorealistic portraits, and commercial product visuals. Simply enter a prompt to produce 1 image output with no file inputs. Ideal for marketing assets, cinematic concept art, and creative projects requiring polished, text-accurate visuals.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.35.1",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-image"
          ]
        }
      },
      {
        "name": "api_openrouter_mai_image_26_flash",
        "title": "Mai Image 2.6 Flash: Text to Image",
        "task": "Text to Image",
        "model": "Mai Image 2.6",
        "freshness": "recent",
        "usage": 16,
        "recommend": "not_recommended",
        "description": "Generate images from text prompts with Microsoft's MAI Image 2.6 Flash, a fast, cost-efficient model that emphasizes low latency and high throughput for rapid visual ideation. The workflow takes a single text prompt and produces one image output. Ideal for concept sketches, marketing mockups, and batch asset drafts where speed matters more than final polish.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.35.1",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-image"
          ]
        }
      },
      {
        "name": "video_bernini_r_image_editing",
        "title": "Bernini-R: Image Edit",
        "task": "Image Edit",
        "model": "Bernini-R",
        "freshness": "current",
        "usage": 48,
        "recommend": "not_recommended",
        "description": "Generate an edited image with matched lighting and view a side-by-side before/after comparison. Ideal for portrait and product relighting, consistent lighting across photo sets, and e-commerce catalog photography.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt",
            "image: Edit region mask"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.25.0",
        "capabilities": {
          "workflow": [
            "image-edit"
          ]
        }
      },
      {
        "name": "api_nano_banana_2_lite_t2i",
        "title": "Nano Banana 2 Lite: Text to Image",
        "task": "Text to Image",
        "model": "Nano Banana 2",
        "freshness": "current",
        "usage": 16,
        "recommend": "not_recommended",
        "description": "Generate images from text prompts using the ultra-fast Gemini 3.1 Flash Image model, optimized for rapid creation and up to 4K resolution. This workflow takes a text description and outputs one or more images, supporting API execution for quick iteration and concept visualization.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.27.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-image"
          ],
          "model_options": {
            "GeminiNanoBanana2V2": [
              "Gemini Nano Banana 2.1",
              "Nano Banana 2 (Gemini 3.1 Flash Image)",
              "Nano Banana 2 Lite"
            ]
          }
        }
      },
      {
        "name": "image_qwen_image_2512_with_2steps_lora",
        "title": "Qwen-Image 2512 Turbo",
        "task": "Text to Image",
        "model": "Qwen-Image 2512",
        "freshness": "established",
        "usage": 40,
        "recommend": "not_recommended",
        "description": "Qwen-Image 2512 Image generation with 2-steps Turbo LoRA. Input your text prompt to quickly produce a result, though some quality is sacrificed for speed.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.49",
        "capabilities": {
          "workflow": [
            "text-to-image",
            "turbo"
          ]
        }
      },
      {
        "name": "api_bria_image_edit",
        "title": "Bria: FIBO 2.0 Image Edit",
        "task": "Image Edit",
        "model": "BRIA",
        "freshness": "new",
        "usage": 231,
        "recommend": "medium",
        "description": "Edit an image with Bria FIBO 2.0 by describing the change in plain language, with an optional mask to confine it to one region. FIBO 2.0 is one model for both editing and generation, trained on structured VGL captions, so it holds everything you did not ask to change, and it returns that structured prompt so the edit can be rebuilt or refined from JSON.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.42.3",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ],
          "model_options": {
            "BriaImageEditNode": [
              "FIBO"
            ]
          }
        }
      },
      {
        "name": "api_bria_guided_remove_background",
        "title": "Bria: Guided Background Removal",
        "task": "Remove Background",
        "model": "BRIA",
        "freshness": "new",
        "usage": 36,
        "recommend": "medium",
        "description": "Remove the background from an image with Bria while keeping whatever a text prompt describes, so one subject can be cut out of a busy scene instead of losing everything that is not background. The original pixels come back at full resolution with a new alpha channel. Ideal for product cutouts, portrait isolation and compositing prep.",
        "io": {
          "inputs": [
            "image: Input image"
          ],
          "outputs": [
            "image: Background removed"
          ]
        },
        "minComfyUIVersion": "0.42.5",
        "capabilities": {
          "workflow": [
            "api",
            "background-removal"
          ]
        }
      },
      {
        "name": "api_bria_product_holding",
        "title": "Bria: Product Holding",
        "task": "Image Edit",
        "model": "BRIA",
        "freshness": "new",
        "usage": 0,
        "recommend": "medium",
        "description": "Place a product into the hands of the person in a photo with Bria. One person photo plus one to three product shots is enough for a lifestyle image in which the face, pose and background stay untouched and the product is held naturally. Ideal for e-commerce and social content that would otherwise need a shoot.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.42.5",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ]
        }
      },
      {
        "name": "api_bria_virtual_try_on",
        "title": "Bria: Virtual Try-On",
        "task": "Image Edit",
        "model": "BRIA",
        "freshness": "new",
        "usage": 97,
        "recommend": "medium",
        "description": "Dress the person in a photo with a garment from a separate product shot using Bria. One person photo plus one to three garments or accessories produces a worn look in which the face, pose and background stay untouched and the garment is refitted to the body. Ideal for fashion catalogs and lookbooks without a photoshoot.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.42.5",
        "capabilities": {
          "workflow": [
            "api",
            "fashion",
            "image-edit"
          ]
        }
      },
      {
        "name": "api_meta_muse_image_image_edit",
        "title": "Muse Image 1.0: Image Edit",
        "task": "Image Edit",
        "model": "Muse Image",
        "freshness": "recent",
        "usage": 15,
        "recommend": "not_recommended",
        "description": "Edit an image with precision using Muse Image 1.0, following natural language instructions to alter specific elements while preserving the rest of the scene. Takes 2 input images and produces 1 edited output image. Ideal for product photography touch-ups, creative style transfers, and iterative visual refinement.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.34.4",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ]
        }
      },
      {
        "name": "template_qwen_image_illustration_lora",
        "title": "Qwen Image: Illustration LoRA",
        "task": "Text to Image",
        "model": "Qwen-Image",
        "freshness": "established",
        "usage": 50,
        "recommend": "low",
        "description": "A t2v workflows for a style LoRA built on Qwen-Image. Produces images spanning cute anime character design, European bande dessinee, indie risograph prints, storybook watercolor, retro shoujo manga, and graphic novel illustration.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.49",
        "capabilities": {
          "workflow": [
            "text-to-image",
            "lora"
          ]
        }
      },
      {
        "name": "image_qwen_image",
        "title": "Qwen-Image: Text to Image",
        "task": "Text to Image",
        "model": "Qwen-Image",
        "freshness": "established",
        "usage": 14,
        "recommend": "not_recommended",
        "description": "Generate images with exceptional multilingual text rendering and editing capabilities using Qwen-Image's 20B MMDiT model.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.4.0",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "image_qwen_image_instantx_controlnet",
        "title": "Qwen-Image InstantX Union ControlNet",
        "task": "Image Generation",
        "model": "Qwen-Image",
        "freshness": "established",
        "usage": 45,
        "recommend": "not_recommended",
        "description": "Generate images with Qwen-Image InstantX ControlNet, supporting canny, soft edge, depth, pose",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.42",
        "capabilities": {
          "workflow": [
            "controlnet"
          ]
        }
      },
      {
        "name": "sd3.5_simple_example",
        "title": "SD3.5 Simple",
        "task": "Text to Image",
        "model": "SD3.5",
        "freshness": "established",
        "usage": 29,
        "recommend": "not_recommended",
        "description": "A simple text-to-image workflow that uses Stability AI's next-gen SD3.5 model to generate an image from a text prompt. It accepts a text description as input and outputs a single image, leveraging the model's enhanced quality for realistic results.",
        "io": {
          "inputs": [
            "text: Text prompt for image generation"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.3.5",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "sdxl_simple_example",
        "title": "SDXL Simple",
        "task": "Text to Image",
        "model": "SDXL",
        "freshness": "established",
        "usage": 77,
        "recommend": "low",
        "description": "Generate high-quality images from text prompts using the SDXL model. The user provides a positive prompt and optional negative prompt; the workflow outputs a detailed, high-resolution image via Stable Diffusion XL. This workflow leverages SDXL's strong aesthetic performance for general-purpose text-to-image generation.",
        "io": {
          "inputs": [
            "text×2: Positive prompt describing the desired image and optional negative prompt for undesired elements"
          ],
          "outputs": [
            "image: Generated image from the SDXL model"
          ]
        },
        "minComfyUIVersion": "0.0.1",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "sdxlturbo_example",
        "title": "SDXL Turbo",
        "task": "Text to Image",
        "model": "SDXL",
        "freshness": "established",
        "usage": 117,
        "recommend": "low",
        "description": "A text-to-image workflow using SDXL Turbo, a high-resolution model with strong aesthetics, to produce images from text prompts. It accepts a text prompt and returns the generated image, optimized for fast generation with high visual quality.",
        "io": {
          "inputs": [
            "text: Text prompt for image generation"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.0.1",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "sd3.5_large_depth",
        "title": "SD3.5 Large Depth",
        "task": "Depth Estimation",
        "model": "SD3.5",
        "freshness": "established",
        "usage": 38,
        "recommend": "not_recommended",
        "description": "Generate images guided by depth information using SD 3.5.",
        "io": {
          "inputs": [
            "image: Input image"
          ],
          "outputs": [
            "image: Depth map"
          ]
        },
        "minComfyUIVersion": "0.3.42",
        "capabilities": {
          "workflow": [
            "controlnet"
          ]
        }
      },
      {
        "name": "sdxl_refiner_prompt_example",
        "title": "SDXL Refiner Prompt",
        "task": "Text to Image",
        "model": "SDXL",
        "freshness": "established",
        "usage": 49,
        "recommend": "not_recommended",
        "description": "This workflow uses Stability AI's SDXL model to generate high-resolution images from a text prompt. It produces a single output image with strong aesthetics and coherence, suitable for general-purpose image generation tasks.",
        "io": {
          "inputs": [
            "text: Text prompt for image generation"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.0.1",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "image_flux.1_fill_dev_OneReward",
        "title": "Flux.1 Dev OneReward",
        "task": "Image Generation",
        "model": "FLUX.1",
        "freshness": "established",
        "usage": 79,
        "recommend": "low",
        "description": "Supports various tasks such as image inpainting, outpainting, and object removal by bytedance-research team",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.42",
        "capabilities": {
          "workflow": [
            "inpainting",
            "outpainting"
          ]
        }
      },
      {
        "name": "image_hidream_o1",
        "title": "HiDream O1 Full: Image generation",
        "task": "Text to Image",
        "model": "HiDream",
        "freshness": "current",
        "usage": 20,
        "recommend": "not_recommended",
        "description": "Input a text prompt and optionally upload reference images. Generate a high-resolution image up to 2048x2048 with text-to-image, editing, or subject-driven personalization.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.21.1",
        "capabilities": {
          "workflow": [
            "text-to-image",
            "image-edit"
          ]
        }
      },
      {
        "name": "image_hidream_o1_dev",
        "title": "HiDream O1 Dev",
        "task": "Image Edit",
        "model": "HiDream",
        "freshness": "current",
        "usage": 24,
        "recommend": "not_recommended",
        "description": "Input a text prompt and optional reference images. Generate a high-resolution image (up to 2048x2048) with support for text-to-image, image editing, and subject-driven personalization.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.21.1",
        "capabilities": {
          "workflow": [
            "image-edit",
            "text-to-image"
          ]
        }
      },
      {
        "name": "hidream_e1_1",
        "title": "HiDream E1.1 Image Editing",
        "task": "Image Edit",
        "model": "HiDream",
        "freshness": "established",
        "usage": 36,
        "recommend": "not_recommended",
        "description": "Edit images with HiDream E1.1—superior image quality and editing accuracy compared to HiDream-E1-Full.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.45",
        "capabilities": {
          "workflow": [
            "image-edit"
          ]
        }
      },
      {
        "name": "hidream_i1_dev",
        "title": "HiDream I1 Dev",
        "task": "Text to Image",
        "model": "HiDream",
        "freshness": "established",
        "usage": 17,
        "recommend": "not_recommended",
        "description": "Generate images with HiDream I1 Dev - Balanced version with 28 inference steps, suitable for medium-range hardware.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.29",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "image_newbieimage_exp0_1-t2i",
        "title": "NewBie Exp0.1: Anime Generation",
        "task": "Text to Image",
        "model": "NewBie",
        "freshness": "established",
        "usage": 38,
        "recommend": "not_recommended",
        "description": "Generate detailed anime-style images with NewBie Exp0.1's Next-DiT architecture. Supports XML structured prompts for better multi-character scenes and attribute control.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.3.42",
        "capabilities": {
          "workflow": [
            "anime",
            "text-to-image"
          ]
        }
      },
      {
        "name": "api_bfl_flux_1_kontext_pro_image",
        "title": "Flux.1 Kontext Pro",
        "task": "Image Edit",
        "model": "FLUX",
        "freshness": "established",
        "usage": 40,
        "recommend": "not_recommended",
        "description": "Edit images with text instructions using Flux.1 Kontext Pro. This workflow accepts an input image and a text prompt describing the desired edit, then returns the edited output. Flux is a fast, highly customizable model, though with less realistic colors and textures. This workflow calls a third-party API.",
        "io": {
          "inputs": [
            "image: Input image to edit",
            "text: Text instructions for the desired edit"
          ],
          "outputs": [
            "image: Edited output image"
          ]
        },
        "minComfyUIVersion": "0.3.40",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ]
        }
      },
      {
        "name": "flux_schnell_full_text_to_image",
        "title": "Flux.1 Schnell Full: Text to Image",
        "task": "Text to Image",
        "model": "FLUX.1",
        "freshness": "established",
        "usage": 48,
        "recommend": "not_recommended",
        "description": "Generate images quickly with Flux.1 Schnell full version. Uses Apache2.0 license, requires only 4 steps to generate images while maintaining good image quality.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.0.4",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "image_chroma_text_to_image",
        "title": "Chroma: Text to Image",
        "task": "Text to Image",
        "model": "Chroma",
        "freshness": "established",
        "usage": 107,
        "recommend": "low",
        "description": "Chroma - enhanced Flux model with improved image quality and better prompt understanding for stunning text-to-image generation.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.3.31",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "flux_redux_model_example",
        "title": "Flux.1 Redux Model",
        "task": "Image Generation",
        "model": "FLUX.1",
        "freshness": "established",
        "usage": 20,
        "recommend": "not_recommended",
        "description": "Generate images by transferring style from reference images using Flux.1 Redux.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.0.4",
        "capabilities": {
          "workflow": [
            "controlnet"
          ]
        }
      },
      {
        "name": "flux_schnell",
        "title": "Flux.1 Schnell FP8",
        "task": "Text to Image",
        "model": "FLUX.1",
        "freshness": "established",
        "usage": 35,
        "recommend": "not_recommended",
        "description": "Quickly generate high-quality images from text prompts using Flux.1 Schnell FP8, requiring only 4 sampling steps for efficient inference. This text-to-image workflow takes a text prompt as input and returns a generated image as output.",
        "io": {
          "inputs": [
            "text: Text prompt for image generation"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.0.4",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "api_bfl_flux_1_kontext_multiple_images_input",
        "title": "Flux.1 Kontext Multiple Image Input",
        "task": "Image Edit",
        "model": "FLUX.1 Kontext",
        "freshness": "established",
        "usage": 40,
        "recommend": "not_recommended",
        "description": "This workflow uses Flux.1 Kontext to edit images by combining up to three reference images with a text instruction for style transfer, character consistency, or stitching. It outputs a single edited image and calls a third-party API for execution.",
        "io": {
          "inputs": [
            "image×3: Reference images to guide the edit",
            "text: Text instructions for the desired edit"
          ],
          "outputs": [
            "image: Edited output image"
          ]
        },
        "minComfyUIVersion": "0.3.40",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ]
        }
      },
      {
        "name": "api_bfl_flux2_max_sofa_swap",
        "title": "FLUX.2 [max]: Object Swap",
        "task": "Image Edit",
        "model": "FLUX",
        "freshness": "established",
        "usage": 17,
        "recommend": "not_recommended",
        "description": "Replace objects in images with unmatched quality using FLUX.2 [max]. Perfect for product photography, furniture swaps, and maintaining scene consistency with highest editing precision.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ],
          "model_options": {
            "Flux2ImageNode": [
              "Flux.2 [pro]",
              "Flux.2 [max]"
            ]
          }
        }
      },
      {
        "name": "api_grok_imagine_image_quality_image_generation",
        "title": "Grok Imagine Image Quality: Generation",
        "task": "Text to Image",
        "model": "Grok",
        "freshness": "established",
        "usage": 52,
        "recommend": "low",
        "description": "Generate high-quality images with enhanced detail and text rendering using Grok's advanced image generation modes via API. This fast xAI model produces a detailed image from a text prompt, executed through a third-party API.",
        "io": {
          "inputs": [
            "text: Text prompt describing the desired image"
          ],
          "outputs": [
            "image: Generated high-quality image with enhanced detail and text rendering"
          ]
        },
        "minComfyUIVersion": "0.11.1",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-image"
          ],
          "model_options": {
            "GrokImageNode": [
              "grok-imagine-image-2.0",
              "grok-imagine-image-quality",
              "grok-imagine-image-pro",
              "grok-imagine-image"
            ]
          }
        }
      },
      {
        "name": "template_rob_realistic_2k_images_quick_variations",
        "title": "Realistic 2k Images - Quick Variations",
        "task": "Image Upscale",
        "model": "Grok",
        "freshness": "established",
        "usage": 36,
        "recommend": "not_recommended",
        "description": "Upload an image to create realistic 2K variations. Grok’s image editing capability generates a new version from the input, then the workflow upscales with Z-image for sharp textures. Requires only a text prompt describing the desired variation.",
        "io": {
          "inputs": [
            "image: Reference image to generate a variation of"
          ],
          "outputs": [
            "image: Upscaled 2K variation of the input image"
          ]
        },
        "minComfyUIVersion": "0.11.1",
        "capabilities": {
          "workflow": [
            "api",
            "image-upscale",
            "text-to-image"
          ],
          "model_options": {
            "GrokImageNode": [
              "grok-imagine-image-2.0",
              "grok-imagine-image-quality",
              "grok-imagine-image-pro",
              "grok-imagine-image"
            ]
          }
        }
      },
      {
        "name": "image_mage_flow_edit_int8",
        "title": "Mage-Flow-Edit: Image Edit",
        "task": "Image Edit",
        "model": "Mage-Flow",
        "freshness": "current",
        "usage": 22,
        "recommend": "not_recommended",
        "description": "Edit images by following natural language instructions with Mage-Flow-Edit, an instruction-based model that handles background replacement, style changes, localized edits, and restoration using up to three reference images. The workflow takes one input image and produces one edited output plus a side-by-side comparison view. Ideal for creative photo retouching, product image styling, and restoring old or damaged photos.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.29.0",
        "capabilities": {
          "workflow": [
            "image-edit",
            "int8"
          ]
        }
      },
      {
        "name": "templates_rob_realistic_2k_images_quick_variations.app",
        "title": "Generate Realistic Variations",
        "task": "Image Upscale",
        "model": "Grok",
        "freshness": "established",
        "usage": 28,
        "recommend": "not_recommended",
        "description": "Upload a reference image and enter a text prompt. Grok automatically captions the image, then generates multiple realistic variations while applying a fast 2x upscale that enhances fine details. The result is a set of high-quality, diverse images ready for further use.",
        "io": {
          "inputs": [
            "image: Reference image to generate variations from"
          ],
          "outputs": [
            "image: Generated image variation with enhanced details"
          ]
        },
        "minComfyUIVersion": "0.11.1",
        "capabilities": {
          "workflow": [
            "api",
            "image-upscale"
          ],
          "model_options": {
            "GrokImageNode": [
              "grok-imagine-image-2.0",
              "grok-imagine-image-quality",
              "grok-imagine-image-pro",
              "grok-imagine-image"
            ]
          }
        }
      },
      {
        "name": "image-qwen_image_edit_2511_lora_inflation",
        "title": "Image Inflation Effect with INFL8 LoRA",
        "task": "Image Edit",
        "model": "Qwen-Image-Edit",
        "freshness": "established",
        "usage": 52,
        "recommend": "low",
        "description": "Upload an image and specify a subject. Generate an inflated version of that subject using the INFL8 LoRA for Qwen Image Edit.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.3.60",
        "capabilities": {
          "workflow": [
            "image-edit"
          ]
        }
      },
      {
        "name": "api_kling_v3_t2i",
        "title": "Kling 3.0: Text to Image",
        "task": "Text to Image",
        "model": "Kling",
        "freshness": "established",
        "usage": 45,
        "recommend": "not_recommended",
        "description": "Generate an image from a text prompt using the Kling 3.0 model via a third-party API. Describe your desired scene in natural language, and the workflow returns a high-quality image matching the prompt. Execution time depends on server-side processing.",
        "io": {
          "inputs": [
            "text: Text prompt describing the desired image"
          ],
          "outputs": [
            "image: Generated image from text prompt"
          ]
        },
        "minComfyUIVersion": "0.3.32",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-image"
          ]
        }
      },
      {
        "name": "image_chrono_edit_14B",
        "title": "ChronoEdit 14B: Image Edit",
        "task": "Image Edit",
        "model": "Wan2.1",
        "freshness": "established",
        "usage": 32,
        "recommend": "not_recommended",
        "description": "Image editing powered by video models' dynamic understanding, creating physically plausible results while preserving character and style consistency.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.4.0",
        "capabilities": {
          "workflow": [
            "image-edit"
          ]
        }
      },
      {
        "name": "image_z_image_int8",
        "title": "Z-Image Int8: Text to Image",
        "task": "Text to Image",
        "model": "Z-Image",
        "freshness": "current",
        "usage": 29,
        "recommend": "not_recommended",
        "description": "Generate high-quality, diverse images from text prompts using the Z-Image Int8 model, which leverages an undistilled transformer for precise prompt adherence and broad stylistic coverage. The workflow takes no file inputs and produces 1 image output through a standard text-to-image pipeline. Ideal for professional creative projects, LoRA training and model development, and scenarios requiring robust negative control and high compositional diversity.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.23.0",
        "capabilities": {
          "workflow": [
            "text-to-image",
            "int8"
          ]
        }
      },
      {
        "name": "image_qwen_Image_2512_controlnet",
        "title": "Qwen-Image 2512: Fun Union ControlNet",
        "task": "Image Generation",
        "model": "Qwen-Image 2512",
        "freshness": "established",
        "usage": 39,
        "recommend": "not_recommended",
        "description": "Upload an image and select a control type from Canny, HED, Depth, Pose, MLSD, Scribble, or Grayscale. Generate a new image guided by the chosen structural condition.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "controlnet"
          ]
        }
      },
      {
        "name": "api_openrouter_mai_image_26_flash_image_edit",
        "title": "MAI Image 2.6 Flash: Image Edit",
        "task": "Image Edit",
        "model": "Mai Image 2.6",
        "freshness": "recent",
        "usage": 10,
        "recommend": "not_recommended",
        "description": "Edit images with MAI Image 2.6 Flash, Microsoft's fast image-to-image model that transforms a single input image into a refined output. Ideal for quick retouching, style adjustments, and rapid concept iteration.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.35.1",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ]
        }
      },
      {
        "name": "api_recraft_v4_style_reference",
        "title": "Recraft V4: Style-Reference Image Generation",
        "task": "Style Reference",
        "model": "Recraft",
        "freshness": "recent",
        "usage": 14,
        "recommend": "not_recommended",
        "description": "Generate images in the style of one or more reference images using Recraft V4 Styles, capturing texture, composition, and rendering without any training or fine-tuning. Feed in multiple reference images to sharpen or broaden the style match, and receive a single styled output. Ideal for brand-consistent asset creation, matching a client's existing campaign look, or adapting styles from any external image source.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.34.0",
        "capabilities": {
          "workflow": [
            "api",
            "style-reference"
          ]
        }
      },
      {
        "name": "template-recraft_create_style",
        "title": "Style Transfer - Recraft Text-to-Image",
        "task": "Style Transfer",
        "model": "Recraft",
        "freshness": "established",
        "usage": 23,
        "recommend": "not_recommended",
        "description": "Upload 1–5 style reference images to define a visual brand style via the Recraft API. The model learns a style ID and generates new, sharp, detailed images matching that style using Standard or Pro modes. You supply references and a text prompt; the system returns a consistent, on‑brand image in seconds.",
        "io": {
          "inputs": [
            "image×5: Style reference images (up to 5) to define the visual brand style"
          ],
          "outputs": [
            "image: Generated image matching the trained style ID"
          ]
        },
        "minComfyUIVersion": "0.12.0",
        "capabilities": {
          "workflow": [
            "api",
            "style-transfer"
          ]
        }
      },
      {
        "name": "api_recraft_style_reference",
        "title": "Recraft Style Reference Generation",
        "task": "Style Reference",
        "model": "Recraft",
        "freshness": "established",
        "usage": 22,
        "recommend": "not_recommended",
        "description": "Use up to four reference images to define a consistent visual style, then provide a text prompt to generate a new image that maintains visual coherence. This workflow executes via a third‑party API, producing sharp, detailed images from Recraft in Standard or Pro quality modes.",
        "io": {
          "inputs": [
            "image×4: Reference images to define a consistent visual style",
            "text: Text prompt for the target image using the created style"
          ],
          "outputs": [
            "image: Generated image consistent with the defined style reference"
          ]
        },
        "minComfyUIVersion": "0.12.0",
        "capabilities": {
          "workflow": [
            "api",
            "style-reference"
          ]
        }
      },
      {
        "name": "api_recraft_image_gen_with_color_control",
        "title": "Recraft: Color Control Image Generation",
        "task": "Text to Image",
        "model": "Recraft",
        "freshness": "established",
        "usage": 15,
        "recommend": "not_recommended",
        "description": "This text-to-image workflow generates images with controlled color palettes using Recraft. It accepts a text prompt and outputs an image with tailored color tones via a third-party API. Recraft produces sharp, detailed outputs with Standard and Pro modes, supporting multiple visual styles.",
        "io": {
          "inputs": [
            "text: Text prompt describing the desired image"
          ],
          "outputs": [
            "image: Generated image with controlled color palette"
          ]
        },
        "minComfyUIVersion": "0.3.34",
        "capabilities": {
          "workflow": [
            "text-to-image",
            "api"
          ]
        }
      },
      {
        "name": "image_boogu_image_0_1_turbo_t2i",
        "title": "Boogu Turbo: Text to Image",
        "task": "Text to Image",
        "model": "Boogu",
        "freshness": "current",
        "usage": 13,
        "recommend": "not_recommended",
        "description": "Generate high-quality images from text prompts using the Boogu Turbo model, a distilled 10B-parameter text-to-image pipeline that produces photorealistic outputs in just 4 steps. The workflow accepts one text prompt and outputs a single 1024x1024 image. Ideal for rapid prototyping of photorealistic concepts, generating marketing visuals, and creating bilingual text-rendered images with strong prompt adherence.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.26.0",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "image_boogu_image_0_1_edit_int8",
        "title": "Boogu image 0.1 Edit Int8",
        "task": "Image Edit",
        "model": "Boogu",
        "freshness": "current",
        "usage": 12,
        "recommend": "not_recommended",
        "description": "Edit images with the Boogu Image Edit model, now optimized with convrot int8 quantization for faster inference and better quality than fp8. This workflow takes 1 input image and produces 1 edited image output. Ideal for quick image edits, batch processing with improved performance, and users wanting to leverage the convrot int8 format's speed and quality advantages.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.26.0",
        "capabilities": {
          "workflow": [
            "image-edit",
            "int8"
          ]
        }
      },
      {
        "name": "image_kandinsky5_t2i",
        "title": "Kandinsky 5.0 Image Lite: Text to Image",
        "task": "Text to Image",
        "model": "Kandinsky",
        "freshness": "established",
        "usage": 14,
        "recommend": "not_recommended",
        "description": "A lightweight 2B model that generates images from English and Russian prompts with high visual quality.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.4.0",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "image_qwen_image_union_control_lora",
        "title": "Qwen-Image Union Control",
        "task": "Text to Image",
        "model": "Qwen-Image",
        "freshness": "established",
        "usage": 18,
        "recommend": "not_recommended",
        "description": "Generate images with precise structural control using Qwen-Image's unified ControlNet LoRA. Supports multiple control types including canny, depth, lineart, softedge, normal, and openpose for diverse creative applications.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.4.0",
        "capabilities": {
          "workflow": [
            "text-to-image",
            "controlnet"
          ]
        }
      },
      {
        "name": "image_qwen_image_controlnet_patch",
        "title": "Qwen-Image ControlNet Model Patch",
        "task": "Text to Image",
        "model": "Qwen-Image",
        "freshness": "established",
        "usage": 28,
        "recommend": "not_recommended",
        "description": "Control image generation using Qwen-Image ControlNet models. Supports canny, depth, and inpainting controls through model patching.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.52",
        "capabilities": {
          "workflow": [
            "text-to-image",
            "controlnet"
          ]
        }
      },
      {
        "name": "hidream_i1_full",
        "title": "HiDream I1 Full",
        "task": "Text to Image",
        "model": "HiDream",
        "freshness": "established",
        "usage": 13,
        "recommend": "not_recommended",
        "description": "Generate images with HiDream I1 Full - Complete version with 50 inference steps for highest quality output.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.29",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "hidream_e1_full",
        "title": "HiDream E1 Image Edit",
        "task": "Image Edit",
        "model": "HiDream",
        "freshness": "established",
        "usage": 19,
        "recommend": "not_recommended",
        "description": "Edit images with HiDream E1 - Professional natural language image editing model.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.29",
        "capabilities": {
          "workflow": [
            "image-edit"
          ]
        }
      },
      {
        "name": "image_omnigen2_t2i",
        "title": "OmniGen2: Text to Image",
        "task": "Text to Image",
        "model": "OmniGen",
        "freshness": "established",
        "usage": 15,
        "recommend": "not_recommended",
        "description": "Generate high-quality images from text prompts using OmniGen2's unified 7B multimodal model with dual-path architecture.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.3.42",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "image_omnigen2_image_edit",
        "title": "OmniGen2 Image Edit",
        "task": "Image Edit",
        "model": "OmniGen",
        "freshness": "established",
        "usage": 10,
        "recommend": "not_recommended",
        "description": "Edit images with natural language instructions using OmniGen2's advanced image editing capabilities and text rendering support.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.3.42",
        "capabilities": {
          "workflow": [
            "image-edit"
          ]
        }
      },
      {
        "name": "api_ideogram_v4_t2i",
        "title": "Ideogram v4: Text to Image (API)",
        "task": "Text to Image",
        "model": "Ideogram",
        "freshness": "current",
        "usage": 18,
        "recommend": "not_recommended",
        "description": "Input a text prompt in structured JSON format. Generate a high-quality image with precise layout control, text rendering, and color palette support.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.24.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-image"
          ],
          "model_options": {
            "GeminiNodeV3": [
              "Gemini 3.8 Flash",
              "Gemini 3.7 Flash",
              "Gemini 3.5 Flash",
              "Gemini 3.1 Pro",
              "Gemini 3.1 Flash-Lite"
            ]
          }
        }
      },
      {
        "name": "api_ideogram_p_image_t2i",
        "title": "Ideogram P-Image: Text to Image",
        "task": "Text to Image",
        "model": "Ideogram",
        "freshness": "current",
        "usage": 15,
        "recommend": "not_recommended",
        "description": "Generate images from text descriptions using Ideogram's fast P-Image model, which delivers quality comparable to leading image models at significantly lower cost and latency. One text prompt is accepted and one image is output, with controls for quality level and resolution. Ideal for rapid iteration in design workflows, high-volume content production, and cost-sensitive A/B testing of visual concepts.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.29.2",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-image"
          ]
        }
      },
      {
        "name": "api_luma_uni_1_image_create",
        "title": "Luma UNI-1: Image Creation",
        "task": "Style Reference",
        "model": "Luma UNI-1",
        "freshness": "established",
        "usage": 15,
        "recommend": "not_recommended",
        "description": "Use Luma UNI‑1 to generate original images from a detailed text prompt and up to nine reference images that guide style, character, color, lighting, or composition. The model reasons through the prompt to produce a cohesive result. This workflow calls a third‑party API, so execution time depends on server response.",
        "io": {
          "inputs": [
            "image×9: Reference images defining style, character, color, lighting, or composition",
            "text: Detailed text prompt describing the desired image"
          ],
          "outputs": [
            "image: Generated original image inspired by reference materials"
          ]
        },
        "minComfyUIVersion": "0.20.3",
        "capabilities": {
          "workflow": [
            "api",
            "style-reference"
          ],
          "model_options": {
            "LumaImageNode2": [
              "uni-1",
              "uni-1-max"
            ]
          }
        }
      },
      {
        "name": "api_luma_uni_1_image_edit",
        "title": "Luma UNI-1: Image Edit",
        "task": "Image Edit",
        "model": "Luma UNI-1",
        "freshness": "established",
        "usage": 12,
        "recommend": "not_recommended",
        "description": "Edit existing images using Luma UNI‑1’s Image Edit mode, which reasons through prompts before generating and preserves composition and core elements. The workflow accepts an input image and a text prompt describing desired adjustments, then outputs an edited image. It supports two model variants (uni-1 and uni-1-max) and executes via third‑party API.",
        "io": {
          "inputs": [
            "image: Input image to edit",
            "text: Text prompt describing the desired edits"
          ],
          "outputs": [
            "image: Edited image with preserved composition and targeted adjustments"
          ]
        },
        "minComfyUIVersion": "0.20.3",
        "capabilities": {
          "workflow": [
            "image-edit",
            "api"
          ],
          "model_options": {
            "LumaImageEditNode2": [
              "uni-1",
              "uni-1-max"
            ]
          }
        }
      },
      {
        "name": "api_kling_omni_image",
        "title": "Kling O1: Image",
        "task": "Character Reference",
        "model": "Kling O1",
        "freshness": "established",
        "usage": 20,
        "recommend": "not_recommended",
        "description": "Generate high-quality images from text prompts using Kling O1, with character reference support for up to 10 reference images to guide style and content. Features flexible aspect ratios and 1K/2K resolution options. This workflow calls a third-party API — execution time depends on server-side response.",
        "io": {
          "inputs": [
            "text: Text prompt describing the desired image"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.4.0",
        "capabilities": {
          "workflow": [
            "api",
            "character-reference",
            "text-to-image"
          ]
        }
      },
      {
        "name": "api_google_nano_banana2_image_edit_continuation",
        "title": "Image Edit",
        "task": "Image Edit",
        "model": "Google",
        "freshness": "recent",
        "usage": 17,
        "recommend": "not_recommended",
        "description": "Start with an image and an empty prompt, then describe the edit you want. Nano Banana 2 generates a modified image while preserving subject consistency and high visual fidelity.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.20.3",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ]
        }
      },
      {
        "name": "api_from_photo_2_miniature",
        "title": "Photo to Blueprint to Model",
        "task": "Image Edit",
        "model": "Gemini3 Pro Image Preview",
        "freshness": "established",
        "usage": 27,
        "recommend": "not_recommended",
        "description": "Transform a real building photo into an architectural blueprint and then into a detailed miniature physical model using Google’s Gemini 3 Pro Image Preview. This image-editing workflow runs via third‑party API and outputs a generated miniature model image, providing a complete architectural visualization pipeline.",
        "io": {
          "inputs": [
            "image: Photo of a building to transform"
          ],
          "outputs": [
            "image: Generated miniature model image"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ],
          "model_options": {
            "GeminiImage2Node": [
              "gemini-3-pro-image-preview",
              "Nano Banana 2 (Gemini 3.1 Flash Image)"
            ]
          }
        }
      },
      {
        "name": "templates-color_illustration",
        "title": "Add Color to Line Art Illustration",
        "task": "Image Edit",
        "model": "Gemini3 Pro Image Preview",
        "freshness": "established",
        "usage": 12,
        "recommend": "not_recommended",
        "description": "This workflow colorizes black-and-white line art using Gemini 3 Pro Image Preview, preserving original strokes while adding full color. It runs via third-party API, requiring no local GPU. An alternative model, Nano Banana 2, is also selectable.",
        "io": {
          "inputs": [
            "image: Black and white line art illustration to colorize"
          ],
          "outputs": [
            "image: Colorized illustration"
          ]
        },
        "minComfyUIVersion": "0.3.71",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ],
          "model_options": {
            "GeminiImage2Node": [
              "gemini-3-pro-image-preview",
              "Nano Banana 2 (Gemini 3.1 Flash Image)"
            ]
          }
        }
      },
      {
        "name": "templates-all_in_one-image_edit_models",
        "title": "1 input and multiple editing model comparison",
        "task": "Image Edit",
        "model": "FLUX.2 Dev",
        "freshness": "established",
        "usage": 15,
        "recommend": "not_recommended",
        "description": "upload one input image and select multiple editing models. generate side-by-side outputs to directly compare model effects and quality.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.33.3",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ]
        }
      },
      {
        "name": "api_meta_muse_image_t2i",
        "title": "Muse Image 1.0: Text to Image",
        "task": "Text to Image",
        "model": "Muse Image",
        "freshness": "recent",
        "usage": 18,
        "recommend": "not_recommended",
        "description": "Generate images from text prompts using Meta's Muse Image 1.0 model, which follows instructions faithfully, edits with precision, and composes from multiple references. The workflow takes no file inputs and produces 1 image output, with a Save Image Advanced node handling the final render. Ideal for creative concept exploration, rapid visual prototyping, and producing polished marketing or social media imagery.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.34.4",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-image"
          ]
        }
      },
      {
        "name": "api_runway_reference_to_image",
        "title": "Runway: Reference to Image",
        "task": "Style Reference",
        "model": "Runway",
        "freshness": "established",
        "usage": 14,
        "recommend": "not_recommended",
        "description": "Generate a new image using Runway’s AI, guided by a reference image for style and composition. The workflow accepts an image and text prompt, then outputs a single generated image. Execution requires a third‑party API call, with response time depending on server availability.",
        "io": {
          "inputs": [
            "image: Reference image to guide style and composition",
            "text: Text prompt describing the desired image"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.3.37",
        "capabilities": {
          "workflow": [
            "api",
            "style-reference"
          ]
        }
      },
      {
        "name": "api_runway_text_to_image",
        "title": "Runway: Text to Image",
        "task": "Text to Image",
        "model": "Runway",
        "freshness": "established",
        "usage": 21,
        "recommend": "not_recommended",
        "description": "Generate a high-quality image from a text prompt using Runway’s text-to-image model. The workflow accepts a text description and outputs the resulting image via a third-party API call, with execution time dependent on server response.",
        "io": {
          "inputs": [
            "text: Text prompt describing the desired image"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.3.37",
        "capabilities": {
          "workflow": [
            "text-to-image",
            "api"
          ]
        }
      },
      {
        "name": "flux_depth_lora_example",
        "title": "Flux.1 Depth Lora",
        "task": "Depth Estimation",
        "model": "FLUX",
        "freshness": "established",
        "usage": 20,
        "recommend": "not_recommended",
        "description": "Generate images guided by depth information using Flux.1 LoRA.",
        "io": {
          "inputs": [
            "image: Input image"
          ],
          "outputs": [
            "image: Depth map"
          ]
        },
        "minComfyUIVersion": "0.3.42",
        "capabilities": {
          "workflow": [
            "controlnet"
          ]
        }
      },
      {
        "name": "api_bfl_flux_1_kontext_max_image",
        "title": "Flux.1 Kontext Max",
        "task": "Image Edit",
        "model": "FLUX.1 Kontext",
        "freshness": "established",
        "usage": 15,
        "recommend": "not_recommended",
        "description": "Edit images using the Flux.1 Kontext Max model via a third-party API. This workflow takes an input image and natural language instructions to perform edits like color changes and style transfer, preserving composition and visual style.",
        "io": {
          "inputs": [
            "image: Input image to edit",
            "text: Natural language edit instructions"
          ],
          "outputs": [
            "image: Edited output image"
          ]
        },
        "minComfyUIVersion": "0.3.40",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ]
        }
      },
      {
        "name": "api_bfl_flux_pro_t2i",
        "title": "Flux.1 [Pro]: Text to Image",
        "task": "Image Edit",
        "model": "FLUX",
        "freshness": "established",
        "usage": 15,
        "recommend": "not_recommended",
        "description": "This workflow generates images from text prompts using the Flux model, known for its speed and customizability, though textures may appear less realistic. It outputs a generated image and relies on a third-party API for execution, making it suitable for rapid generation and LoRA integration.",
        "io": {
          "inputs": [
            "text: Text prompt describing the desired image"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.3.32",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ]
        }
      },
      {
        "name": "Image_capybara_v0_1_image_edit",
        "title": "Capybara: Image Edit",
        "task": "Image Edit",
        "model": "Capybara",
        "freshness": "established",
        "usage": 11,
        "recommend": "not_recommended",
        "description": "Upload an image or video and input a text instruction. Generate an edited output using the Capybara model for tasks like style changes, object replacement, or time-of-day adjustments.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "image-edit"
          ]
        }
      },
      {
        "name": "Image_capybara_v0_1_text_to_image",
        "title": "Capybara: Text to Image",
        "task": "Text to Image",
        "model": "Capybara",
        "freshness": "established",
        "usage": 14,
        "recommend": "not_recommended",
        "description": "Input a text prompt to generate an image. This workflow uses the Capybara unified model for text-to-image synthesis.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.3.71",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "image_chroma1_radiance_text_to_image",
        "title": "Chroma1 Radiance Text to Image",
        "task": "Text to Image",
        "model": "Chroma",
        "freshness": "established",
        "usage": 19,
        "recommend": "not_recommended",
        "description": "Chroma1-Radiance works directly with image pixels instead of compressed latents, delivering higher quality images with reduced artifacts and distortion.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.3.60",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "api_bytedance_seedream_5_0_lite_t2i",
        "title": "Seedream 5.0 Lite: Text to Image",
        "task": "Text to Image",
        "model": "Seedream 5.0",
        "freshness": "established",
        "usage": 83,
        "recommend": "low",
        "description": "Input your text prompt to generate a detailed image. This workflow creates a visual representation of your description with precise control over style and layout.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.31.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-image"
          ],
          "model_options": {
            "ByteDanceSeedreamNodeV3": [
              "seedream 5.0 pro",
              "seedream 5.0 flash",
              "seedream 5.0 lite",
              "seedream-4-5-251128",
              "seedream-4-0-250828"
            ]
          }
        }
      },
      {
        "name": "image_ernie_image_turbo",
        "title": "Ernie Image Turbo: Text To Image",
        "task": "Text to Image",
        "model": "ERNIE-Image",
        "freshness": "established",
        "usage": 28,
        "recommend": "not_recommended",
        "description": "Generate images from text prompts using the ERNIE-Image turbo model. Input a text description and receive a high-quality image with precise text rendering and structured layouts.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.15.0",
        "capabilities": {
          "workflow": [
            "text-to-image",
            "turbo"
          ]
        }
      },
      {
        "name": "image_ernie_image",
        "title": "Ernie Image: Text to Image",
        "task": "Text to Image",
        "model": "ERNIE-Image",
        "freshness": "established",
        "usage": 54,
        "recommend": "low",
        "description": "Generate images from text prompts using the ERNIE-Image model. Input a text description to produce detailed, structured visuals with a broad stylistic range.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.15.0",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "sd3.5_large_canny_controlnet_example",
        "title": "SD3.5 Large Canny ControlNet",
        "task": "Image Generation",
        "model": "SD3.5",
        "freshness": "established",
        "usage": 209,
        "recommend": "medium",
        "description": "Generate images guided by edge detection using SD 3.5 Canny ControlNet.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.5",
        "capabilities": {
          "workflow": [
            "controlnet"
          ]
        }
      },
      {
        "name": "sd3.5_large_blur",
        "title": "SD3.5 Large Blur",
        "task": "Image Edit",
        "model": "SD3.5",
        "freshness": "established",
        "usage": 9,
        "recommend": "not_recommended",
        "description": "Generate images guided by blurred reference images using SD 3.5.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.5",
        "capabilities": {
          "workflow": [
            "image-edit"
          ]
        }
      },
      {
        "name": "sdxl_revision_text_prompts",
        "title": "SDXL Revision Text Prompts",
        "task": "Text to Image",
        "model": "SDXL",
        "freshness": "established",
        "usage": 19,
        "recommend": "not_recommended",
        "description": "Generate images by transferring concepts from reference images using SDXL Revision.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.0.1",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "api_luma_photon_i2i",
        "title": "Luma Photon: Image to Image",
        "task": "Image Edit",
        "model": "Luma",
        "freshness": "established",
        "usage": 14,
        "recommend": "not_recommended",
        "description": "This workflow uses Luma’s legacy Photon model (Photon Flash 1 for speed or Photon 1 for quality) to generate an image from a reference image and a text prompt. It relies on a third‑party API, so run time depends on server response. The model is an older medium model with natural motion capabilities.",
        "io": {
          "inputs": [
            "image: Reference image to guide generation",
            "text: Text prompt describing the desired image"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.3.32",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ]
        }
      },
      {
        "name": "api_luma_photon_style_ref",
        "title": "Luma Photon: Style Reference",
        "task": "Text to Image",
        "model": "Luma",
        "freshness": "established",
        "usage": 9,
        "recommend": "not_recommended",
        "description": "Generate images by blending up to two style reference images with a text prompt using Luma Photon API. The output adopts the aesthetic from the references while matching the prompt. This workflow uses a third-party API, so execution time depends on server-side response.",
        "io": {
          "inputs": [
            "image×2: Style reference images to guide the aesthetic",
            "text: Text prompt describing the desired image"
          ],
          "outputs": [
            "image: Generated output image"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-image"
          ]
        }
      },
      {
        "name": "api_magnific_image_style_transfer",
        "title": "Magnific: Style Reference",
        "task": "Style Reference",
        "model": "Magnific",
        "freshness": "established",
        "usage": 24,
        "recommend": "not_recommended",
        "description": "This style reference workflow uses Magnific to merge the structure of a target image with the artistic style of a reference image. Upload both images, and it generates a new image combining content and style. The workflow calls a third-party API, so execution time depends on server response.",
        "io": {
          "inputs": [
            "image: Target image providing the structure and content",
            "image: Style reference image providing the artistic style"
          ],
          "outputs": [
            "image: Image with target structure rendered in reference style"
          ]
        },
        "minComfyUIVersion": "0.11.0",
        "capabilities": {
          "workflow": [
            "api",
            "style-reference"
          ]
        }
      },
      {
        "name": "image_pixeldit_t2i",
        "title": "PixelDiT: Text to Image",
        "task": "Text to Image",
        "model": "PixelDiT",
        "freshness": "established",
        "usage": 20,
        "recommend": "not_recommended",
        "description": "Input a text prompt and optional negative prompt. Generate a 1024px image using PixelDiT's VAE-free pixel diffusion transformer.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.16.0",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "image_longcat_text_to_image",
        "title": "LongCat Image: Text to Image",
        "task": "Text to Image",
        "model": "LongCat",
        "freshness": "established",
        "usage": 10,
        "recommend": "not_recommended",
        "description": "Generate an image from a text prompt. Input your desired scene description in English or Chinese. Output a photorealistic image with accurate text rendering.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.16.0",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "api_wan_text_to_image",
        "title": "Wan2.5: Text to Image",
        "task": "Text to Image",
        "model": "Wan2.5",
        "freshness": "established",
        "usage": 19,
        "recommend": "not_recommended",
        "description": "Generate images from text prompts using Wan2.5 (wan2.5-t2i-preview), offering strong prompt adherence and visual quality. This workflow relies on a third-party API, so execution time depends on server-side response. The only input is a text description; the output is the generated image.",
        "io": {
          "inputs": [
            "text: Text prompt describing the desired image"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.3.60",
        "capabilities": {
          "workflow": [
            "text-to-image",
            "api"
          ],
          "model_options": {
            "WanTextToImageApi": [
              "wan2.5-t2i-preview"
            ]
          }
        }
      },
      {
        "name": "template_qwen_Image_2512_360_lora",
        "title": "Qwen Image 2512: 360 Panorama Image",
        "task": "Text to Image",
        "model": "Qwen-Image 2512",
        "freshness": "current",
        "usage": 25,
        "recommend": "not_recommended",
        "description": "Generate 360-degree equirectangular projection images from text descriptions using a rank 128 LoRA on the 20B MMDiT model.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.4.0",
        "capabilities": {
          "workflow": [
            "text-to-image",
            "lora"
          ]
        }
      },
      {
        "name": "api_recraft_v4_t2i",
        "title": "Recraft v4: Text to Image",
        "task": "Text to Image",
        "model": "Recraft",
        "freshness": "established",
        "usage": 9,
        "recommend": "not_recommended",
        "description": "Generate high-resolution images from a text prompt using Recraft’s V4 model, delivering sharp, detailed outputs with enhanced realism. Choose between the standard `recraftv4` (1024x1024) or `recraftv4_pro` (2048x2048) variant. This workflow calls a third-party API, so response time depends on server availability.",
        "io": {
          "inputs": [
            "text: Text prompt describing the desired image"
          ],
          "outputs": [
            "image: Generated high-resolution image with enhanced realism and detail"
          ]
        },
        "minComfyUIVersion": "0.14.1",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-image"
          ],
          "model_options": {
            "RecraftV4TextToImageNode": [
              "recraftv4_1",
              "recraftv4_1_flash",
              "recraftv4_1_utility",
              "recraftv4_1_pro",
              "recraftv4_1_utility_pro",
              "recraftv4",
              "recraftv4_pro",
              "recraftv4_styles",
              "recraftv4_styles_pro"
            ]
          }
        }
      },
      {
        "name": "api_bria_image_outpainting",
        "title": "Bria: Image Outpainting",
        "task": "Image Generation",
        "model": "BRIA",
        "freshness": "established",
        "usage": 34,
        "recommend": "not_recommended",
        "description": "Extend images beyond their original boundaries using BRIA's outpainting workflow. The input image is padded and seamlessly expanded via edge-aware scene generation with the FIBO model for structured, text-guided editing. This template relies on a third-party API, so completion time depends on server-side response.",
        "io": {
          "inputs": [
            "image: Input image to expand outward"
          ],
          "outputs": [
            "image: Outpainted image with extended boundaries"
          ]
        },
        "minComfyUIVersion": "0.10.0",
        "capabilities": {
          "workflow": [
            "api",
            "outpainting"
          ],
          "model_options": {
            "BriaImageEditNode": [
              "FIBO"
            ]
          }
        }
      },
      {
        "name": "hidream_i1_fast",
        "title": "HiDream I1 Fast",
        "task": "Text to Image",
        "model": "HiDream",
        "freshness": "established",
        "usage": 8,
        "recommend": "not_recommended",
        "description": "Generate images quickly with HiDream I1 Fast - Lightweight version with 16 inference steps, ideal for rapid previews on lower-end hardware.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.29",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "api_openai_fashion_billboard_generator",
        "title": "Fashion Billboard Generator",
        "task": "Image Edit",
        "model": "GPT-Image-1.5",
        "freshness": "established",
        "usage": 9,
        "recommend": "not_recommended",
        "description": "Create professional mall billboard advertisements with realistic fashion models from a text prompt. This API workflow leverages GPT-Image-1.5’s exceptional prompt adherence and image editing strengths, including virtual try-on and product mockup capabilities, to output a polished billboard advertisement.",
        "io": {
          "inputs": [
            "image: Clothing or outfit photo to use in the advertisement"
          ],
          "outputs": [
            "image: Generated billboard advertisement image"
          ]
        },
        "minComfyUIVersion": "0.21.1",
        "capabilities": {
          "workflow": [
            "image-edit",
            "api"
          ],
          "model_options": {
            "OpenAIGPTImageNodeV2": [
              "gpt-image-2.5-flare",
              "gpt-image-2.5-sunburst",
              "gpt-image-2",
              "gpt-image-1.5",
              "gpt-image-1"
            ]
          }
        }
      },
      {
        "name": "api_grok_imagine_image_quality_image_edit",
        "title": "Grok Imagine Image Quality: Edit",
        "task": "Image Edit",
        "model": "Grok",
        "freshness": "established",
        "usage": 8,
        "recommend": "not_recommended",
        "description": "Edit images with enhanced quality using Grok Imagine Image Quality mode, leveraging Grok's fast performance. Provide an input image and a text prompt describing the desired edit to improve detail and creative control. This workflow calls a third-party API for execution.",
        "io": {
          "inputs": [
            "image: Input image to edit",
            "text: Text prompt describing the desired edit"
          ],
          "outputs": [
            "image: Edited image with enhanced quality and fidelity"
          ]
        },
        "minComfyUIVersion": "0.11.1",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ],
          "model_options": {
            "GrokImageNode": [
              "grok-imagine-image-2.0",
              "grok-imagine-image-quality",
              "grok-imagine-image-pro",
              "grok-imagine-image"
            ]
          }
        }
      },
      {
        "name": "templates-1_input-multiple_styles_prompt.app",
        "title": "1 input multiple styles from prompt",
        "task": "Image Edit",
        "model": "Grok",
        "freshness": "established",
        "usage": 4,
        "recommend": "not_recommended",
        "description": "Upload an image and a text prompt to Grok's image edit API. In seconds, it produces 12 stylized variations of your input, leveraging fast inference and high-quality output. This workflow demonstrates versatile image editing from a single prompt.",
        "io": {
          "inputs": [
            "image: Source image to generate multiple stylized variations of"
          ],
          "outputs": [
            "image×12: Stylized variations of the input image in different styles"
          ]
        },
        "minComfyUIVersion": "0.11.1",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ],
          "model_options": {
            "GrokImageEditNodeV2": [
              "grok-imagine-image-2.0",
              "grok-imagine-image-quality",
              "grok-imagine-image-pro",
              "grok-imagine-image"
            ]
          }
        }
      },
      {
        "name": "image_ovis_text_to_image",
        "title": "Ovis-Image Text to Image",
        "task": "Text to Image",
        "model": "Ovis-Image",
        "freshness": "established",
        "usage": 8,
        "recommend": "not_recommended",
        "description": "Ovis-Image is a 7B text-to-image model specifically optimized for high-quality text rendering in generated images. Designed to operate efficiently under computational constraints, it excels at accurately generating images containing text content.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.4.0",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "template_sugar_coated_gummy_style_qwen",
        "title": "Qwen Image: Gummy Animals LoRA",
        "task": "Text to Image",
        "model": "Qwen-Image",
        "freshness": "established",
        "usage": 12,
        "recommend": "not_recommended",
        "description": "Apply a Qwen Image LoRA to transform any subject into a sour gummy candy-style character. Enter an animal name in the prompt to generate a stylized, edible-looking creature.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.42",
        "capabilities": {
          "workflow": [
            "text-to-image",
            "lora"
          ]
        }
      },
      {
        "name": "api_recraft_image_gen_with_style_control",
        "title": "Recraft: Style Control Image Generation",
        "task": "Text to Image",
        "model": "Recraft",
        "freshness": "established",
        "usage": 2,
        "recommend": "not_recommended",
        "description": "This text‑to‑image workflow uses Recraft’s API to generate sharp, detailed images with precise style control, applying visual examples, positioning alignment, and object fine‑tuning. You input a text prompt and receive a generated image; execution time depends on the server‑side API response.",
        "io": {
          "inputs": [
            "text: Text prompt describing the desired image"
          ],
          "outputs": [
            "image: Generated image with style control applied"
          ]
        },
        "minComfyUIVersion": "0.3.34",
        "capabilities": {
          "workflow": [
            "text-to-image",
            "api"
          ]
        }
      },
      {
        "name": "image_ming_image_01_design_image_edit",
        "title": "Ming Image 0.1 Design: Image Edit",
        "task": "Image Edit",
        "model": "Ming Image",
        "freshness": "recent",
        "usage": 31,
        "recommend": "not_recommended",
        "description": "Edit an existing image with Ming Image 0.1 Design, a 6B model tuned for UI, infographics, and posters. It takes one input image plus the boxes you draw on it — position, description, and colour palette — and rebuilds them into a layer-by-layer caption, returning one re-rendered image. Ideal for restaging a layout, poster-style re-renders, and turning a photograph into a designed composition.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt",
            "image: Edit region mask"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.38.0",
        "capabilities": {
          "workflow": [
            "image-edit"
          ]
        }
      },
      {
        "name": "image_ming_image_01_design_t2i",
        "title": "Ming Image 0.1 Design: Text to Image",
        "task": "Text to Image",
        "model": "Ming Image",
        "freshness": "recent",
        "usage": 20,
        "recommend": "not_recommended",
        "description": "Generate text-heavy design compositions from text prompts using Ming Image 0.1 Design, a 6B model tuned for UI screens, infographics, and posters. It takes no file inputs and produces one image, supporting transparent backgrounds for easy layering. Ideal for interface mockups, marketing posters, and data-rich infographic drafts.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated image"
          ]
        },
        "minComfyUIVersion": "0.37.3",
        "capabilities": {
          "workflow": [
            "int8",
            "text-to-image"
          ]
        }
      },
      {
        "name": "image_anima_lllite_any_control_to_image",
        "title": "Anima Lllite: Any Control to Image",
        "task": "Image Generation",
        "model": "Anima",
        "freshness": "current",
        "usage": 38,
        "recommend": "not_recommended",
        "description": "Apply lightweight, LoRA-like conditional control to Anima-Base v1.0 anime-style images using ControlNet-LLLite, accepting one conditioning input and one optional mask for inpainting to produce a single generated output. Ideal for anime inpainting with dynamic masking, style transfer from lineart or grayscale sketches, and controlled character generation in multi-scene compositions.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.29.0",
        "capabilities": {
          "workflow": [
            "anime",
            "controlnet"
          ]
        }
      },
      {
        "name": "image_anima_lllite_image_inpainting",
        "title": "Anima Lllite: Image Inpainting",
        "task": "Image Inpainting",
        "model": "Anima",
        "freshness": "current",
        "usage": 64,
        "recommend": "low",
        "description": "Inpaint anime images with Anima LLLite, a lightweight ControlNet variant that applies low-rank corrections for precise mask-based editing. Input 1 image and receive 1 inpainted output plus a comparison view. Ideal for fixing character details, removing unwanted objects from anime scenes, and seamless background restoration.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.29.0",
        "capabilities": {
          "workflow": [
            "anime",
            "inpainting"
          ]
        }
      },
      {
        "name": "image_anima_lllite_depth_control_to_image",
        "title": "Anima Lllite: Depth Control to Image",
        "task": "Depth Estimation",
        "model": "Anima",
        "freshness": "current",
        "usage": 0,
        "recommend": "not_recommended",
        "description": "Generate anime-style illustrations from a depth map using the Anima LLLite model, which applies a lightweight ControlNet correction for precise spatial control. Input 1 image to produce 2 image outputs plus a comparison view. Ideal for 3D scene reconstruction, VFX depth passes, and AR asset preparation.",
        "io": {
          "inputs": [
            "image: Input image"
          ],
          "outputs": [
            "image: Depth map"
          ]
        },
        "minComfyUIVersion": "0.29.0",
        "capabilities": {
          "workflow": [
            "anime"
          ]
        }
      },
      {
        "name": "image_joyai_image_edit",
        "title": "JoyAI Image Edit",
        "task": "Image Edit",
        "model": "JoyAI Image Edit",
        "freshness": "current",
        "usage": 12,
        "recommend": "not_recommended",
        "description": "Edit images by following natural language instructions with JoyAI-Image-Edit, which leverages multimodal understanding for precise, spatially aware modifications. This workflow takes one reference image and produces one edited image output plus a side-by-side comparison view. Ideal for instruction-guided retouching, region-specific object replacement, and complex scene adjustments requiring relational grounding.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Edit prompt"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.29.0",
        "capabilities": {
          "workflow": [
            "image-edit",
            "int8"
          ]
        }
      }
    ]
  },
  {
    "category": "Image Tools",
    "description": "Tooling workflows for image processing helpers such as upscaling, background removal, inpainting, outpainting, relighting, pose and depth extraction, and layer decomposition.",
    "templates": [
      {
        "name": "utility_seedvr2_7b_int8_upscale_image",
        "title": "Image Upscale: SeedVR2 7B Int8",
        "task": "Image Upscale",
        "model": "SeedVR2",
        "freshness": "current",
        "usage": 646,
        "recommend": "high",
        "description": "Upscale images using SeedVR2 7B Int8, a one-step diffusion model that enhances resolution through adversarial training and adaptive window attention. Input 1 image and receive 1 upscaled output with improved detail and clarity. Ideal for restoring low-resolution photos, preparing assets for print, or sharpening lightly degraded digital images.",
        "io": {
          "inputs": [
            "image: Input image"
          ],
          "outputs": [
            "image: Upscaled image"
          ]
        },
        "minComfyUIVersion": "0.28.0",
        "capabilities": {
          "workflow": [
            "image-upscale",
            "int8"
          ]
        }
      },
      {
        "name": "template_qwen_multiangle.app",
        "title": "Generate Multiple Angles with 3D Camera Node",
        "task": "Image Edit",
        "model": "Qwen-Image-Edit",
        "freshness": "new",
        "usage": 928,
        "recommend": "high",
        "description": "Generate camera angle variations of your input with an easy to use 3D camera angle node.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.40.0",
        "capabilities": {
          "workflow": [
            "image-edit"
          ]
        }
      },
      {
        "name": "utility_z_image_turbo_2k_upscaler.app",
        "title": "Image Upscale: Z-Image-Turbo 2K",
        "task": "Image Upscale",
        "model": "Z-Image-Turbo",
        "freshness": "established",
        "usage": 308,
        "recommend": "medium",
        "description": "Upload an image to upscale it to 2K resolution using the Z-Image-Turbo model.",
        "io": {
          "inputs": [
            "image: Input image"
          ],
          "outputs": [
            "image: Upscaled image"
          ]
        },
        "minComfyUIVersion": "0.9.0",
        "capabilities": {
          "workflow": [
            "image-upscale"
          ]
        }
      },
      {
        "name": "image_marigold_v2_depth_estimation",
        "title": "Depth Estimation: Marigold V2",
        "task": "Depth Estimation",
        "model": "Marigold v2",
        "freshness": "recent",
        "usage": 118,
        "recommend": "low",
        "description": "Estimate scene depth from an input image using Marigold V2, generating a depth map with brighter areas representing closer objects and darker areas representing farther objects.",
        "io": {
          "inputs": [
            "image: Input image"
          ],
          "outputs": [
            "image: Depth map"
          ]
        },
        "minComfyUIVersion": "0.36.0",
        "capabilities": {
          "workflow": [
            "depth",
            "int8"
          ]
        }
      },
      {
        "name": "api_topaz_image_enhance_wonder3_5",
        "title": "Image Upscale: Topaz Wonder 3.5",
        "task": "Image Upscale",
        "model": "Topaz",
        "freshness": "current",
        "usage": 227,
        "recommend": "medium",
        "description": "Upscale and enhance a single image using Topaz's Wonder 3.5 generative model, which adds fine detail, reduces noise, and sharpens text and structured graphics while offering Low, Medium, and High enhancement strengths. The workflow takes 1 input image and produces 1 enhanced output alongside a side-by-side comparison view. Ideal for restoring old or compressed photos, improving product shots and marketing visuals, and preparing low-resolution images for print or large-format display.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.30.2",
        "capabilities": {
          "workflow": [
            "api",
            "image-upscale"
          ],
          "model_options": {
            "TopazImageEnhanceV2": [
              "Reimagine",
              "Bloom 2",
              "Wonder 3.5"
            ]
          }
        }
      },
      {
        "name": "api_topaz_image_enhance_wonder4",
        "title": "Image Upscale: Topaz Wonder 4",
        "task": "Image Upscale",
        "model": "Topaz",
        "freshness": "new",
        "usage": 0,
        "recommend": "medium",
        "description": "Upscale a soft, compressed image with Topaz's Wonder 4 generative model: it rebuilds plausible detail instead of stretching pixels, keeping torn edges, seams and flat color fields intact at 1x to 6x. The workflow takes 1 input image and produces 1 enhanced output next to a side-by-side comparison. Ideal for collage, illustration and graphic work that needs to survive a small or compressed source.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.44.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-upscale"
          ],
          "model_options": {
            "TopazImageEnhanceV2": [
              "Reimagine",
              "Bloom 2",
              "Wonder 3.5"
            ]
          }
        }
      },
      {
        "name": "api_magnific_image_upscale_creative",
        "title": "Image Upscale: Magnific Creative",
        "task": "Image Upscale",
        "model": "Magnific",
        "freshness": "established",
        "usage": 323,
        "recommend": "medium",
        "description": "Use this workflow to creatively upscale an image up to 16×, adding detail and texture through the Magnific model’s enhancement capabilities. It accepts an input image and outputs a high‑resolution version with improved clarity and style. This workflow executes via a third‑party API, so response time depends on server availability.",
        "io": {
          "inputs": [
            "image: Input image to upscale creatively"
          ],
          "outputs": [
            "image: Output creatively upscaled image with enhanced detail"
          ]
        },
        "minComfyUIVersion": "0.11.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-upscale"
          ]
        }
      },
      {
        "name": "api_magnific_skin_enhancer",
        "title": "Portrait Enhancement: Magnific",
        "task": "Image Generation",
        "model": "Magnific",
        "freshness": "established",
        "usage": 147,
        "recommend": "low",
        "description": "This workflow enhances skin in portrait images using the Magnific AI model, adjusting lighting, color grading, and realism for improved textures and overall quality. It calls a third-party API for processing, with execution time depending on server-side response.",
        "io": {
          "inputs": [
            "image: Input portrait image for skin enhancement"
          ],
          "outputs": [
            "image: Output portrait with enhanced skin textures and improved lighting"
          ]
        },
        "minComfyUIVersion": "0.11.0",
        "capabilities": {
          "workflow": [
            "api",
            "portrait"
          ]
        }
      },
      {
        "name": "utility_nanobanana_pro_ai_image_fix",
        "title": "Image Upscale: Nano Banana Pro",
        "task": "Image Upscale",
        "model": "Google",
        "freshness": "established",
        "usage": 223,
        "recommend": "medium",
        "description": "Upload an AI-generated image to upscale and enhance details using the Nano Banana Pro model, powered by Google Gemini. This workflow generates a 4K output with improved sharpness and coherence, helping fix common AI generation artifacts. Execution relies on a third-party API, so completion time depends on server-side response.",
        "io": {
          "inputs": [
            "image: Input AI-generated image to enhance and upscale"
          ],
          "outputs": [
            "image: Output upscaled 4K image with improved sharpness and coherence"
          ]
        },
        "minComfyUIVersion": "0.3.71",
        "capabilities": {
          "workflow": [
            "api",
            "image-upscale"
          ],
          "model_options": {
            "GeminiImage2Node": [
              "gemini-3-pro-image-preview",
              "Nano Banana 2 (Gemini 3.1 Flash Image)"
            ]
          }
        }
      },
      {
        "name": "image_qwen_image_2_1_background_removal",
        "title": "Remove Background: Qwen Image 2.1",
        "task": "Remove Background",
        "model": "Qwen Image 2.1",
        "freshness": "recent",
        "usage": 239,
        "recommend": "medium",
        "description": "Remove image backgrounds with Qwen Image 2.1, which natively supports 2K output and RGBA transparency for direct transparent results. It takes 1 input image and returns 1 background-free image plus a comparison view showing the before and after. Ideal for e-commerce product photos, profile pictures, and marketing graphics needing clean cutouts.",
        "io": {
          "inputs": [
            "image: Input image"
          ],
          "outputs": [
            "image: Background removed"
          ]
        },
        "minComfyUIVersion": "0.37.0",
        "capabilities": {
          "workflow": [
            "background-removal",
            "int8"
          ]
        }
      },
      {
        "name": "utility_seedvr2_3b_int8_upscale_image",
        "title": "Image Upscale: SeedVR2 3B Int8",
        "task": "Image Upscale",
        "model": "SeedVR2",
        "freshness": "current",
        "usage": 247,
        "recommend": "medium",
        "description": "Upscale images using SeedVR2 3B Int8, a one-step diffusion-based video restoration model that produces high-quality results with improved temporal consistency. Ideal for restoring degraded footage, upscaling low-resolution videos, and enhancing archival media.",
        "io": {
          "inputs": [
            "image: Input image"
          ],
          "outputs": [
            "image: Upscaled image"
          ]
        },
        "minComfyUIVersion": "0.28.0",
        "capabilities": {
          "workflow": [
            "image-upscale",
            "int8"
          ]
        }
      },
      {
        "name": "utility_image_segment_sam3",
        "title": "Image Segmentation: SAM3",
        "task": "Image Segmentation",
        "model": "SAM3",
        "freshness": "established",
        "usage": 127,
        "recommend": "low",
        "description": "Use the SAM3 model to segment the main subject or content from a photo or image, isolating specific objects or regions. Input an image and receive a segmented mask output.",
        "io": {
          "inputs": [
            "image: Input image"
          ],
          "outputs": [
            "image: Segmentation mask"
          ]
        },
        "minComfyUIVersion": "0.20.0",
        "capabilities": {
          "workflow": [
            "element-segmentation"
          ]
        }
      },
      {
        "name": "utility_moge3_geometry_estimation",
        "title": "Geometry Estimation: MoGe 3",
        "task": "Depth Estimation",
        "model": "MoGe",
        "freshness": "recent",
        "usage": 76,
        "recommend": "low",
        "description": "Estimate fine-detailed geometry from a single image with MoGe 3, the sparse-volumetric refinement version of MoGe. It lifts a coarse point map onto a sparse voxel shell and refines depth with lightweight sparse 3D convolutions, keeping thin structures and depth discontinuities that image-space decoders blur. Ships with the ViT-L checkpoint; the ViT-G file is selectable in the same Load MoGe Model dropdown for higher fine-detail accuracy. Takes 1 image input and outputs 1 depth map and 1 normal map.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.37.0",
        "capabilities": {
          "workflow": [
            "depth-estimation",
            "normal-estimation"
          ]
        }
      },
      {
        "name": "image_qwen_image_layered",
        "title": "Layer Decomposition: Qwen-Image-Layered",
        "task": "Image Generation",
        "model": "Qwen-Image-Layered",
        "freshness": "established",
        "usage": 70,
        "recommend": "low",
        "description": "Decompose an image into editable RGBA layers for high-fidelity recolor, replace, resize, and reposition workflows using Qwen-Image-Layered.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.6.0",
        "capabilities": {
          "workflow": [
            "layer-decompose"
          ]
        }
      },
      {
        "name": "utility_birefnet_remove_background",
        "title": "Remove Background: BiRefNet",
        "task": "Remove Background",
        "model": "BiRefNet",
        "freshness": "current",
        "usage": 107,
        "recommend": "low",
        "description": "Upload an image with any background. Generate a version with the background removed and a precision segmentation mask.",
        "io": {
          "inputs": [
            "image: Input image"
          ],
          "outputs": [
            "image: Background removed"
          ]
        },
        "minComfyUIVersion": "0.21.0",
        "capabilities": {
          "workflow": [
            "background-removal"
          ]
        }
      },
      {
        "name": "utility_depth_anything3_image_depth_estimation",
        "title": "Depth Estimation: Depth Anything 3",
        "task": "Depth Estimation",
        "model": "Depth Anything v3",
        "freshness": "current",
        "usage": 20,
        "recommend": "not_recommended",
        "description": "Upload 1 image. Generate a depth map using Depth Anything 3 and view a side-by-side comparison of the original and depth output. Ideal for 3D scene reconstruction, AR/VR asset preparation, and visual effects depth passes.",
        "io": {
          "inputs": [
            "image: Input image"
          ],
          "outputs": [
            "image: Depth map"
          ]
        },
        "minComfyUIVersion": "0.25.0",
        "capabilities": {
          "workflow": [
            "controlnet"
          ]
        }
      },
      {
        "name": "api_bria_genfill",
        "title": "Inpainting: Bria Generative Fill",
        "task": "Image Generation",
        "model": "BRIA",
        "freshness": "recent",
        "usage": 39,
        "recommend": "not_recommended",
        "description": "Use Bria's Generative Fill API to regenerate a specific region of an image while keeping everything outside the mask pixel-perfect unchanged. Paint a mask over the target area, write a detailed prompt for the replacement content, and compare the before/after side-by-side before exporting. Ideal for product photography corrections, object removal or replacement, and creative compositing where precise control over the edited region matters.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.33.0",
        "capabilities": {
          "workflow": [
            "api",
            "inpainting"
          ]
        }
      },
      {
        "name": "api_bria_expand_image",
        "title": "Image Outpainting: Bria",
        "task": "Image Generation",
        "model": "BRIA",
        "freshness": "recent",
        "usage": 91,
        "recommend": "low",
        "description": "Expand an image beyond its original borders using Bria's outpainting API, generating new surrounding content that matches the original lighting, perspective, and texture while leaving the source untouched. Choose from preset aspect ratios, custom dimensions, or manual canvas placement, with optional text prompts to guide the expansion direction or an auto-generated prompt when left blank. Ideal for extending compositions for print layouts, creating panoramic versions of existing photos, and adapting artwork to new formats or social media dimensions.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.33.0",
        "capabilities": {
          "workflow": [
            "api",
            "outpainting"
          ]
        }
      },
      {
        "name": "utility_bria_remove_image_background",
        "title": "Remove Background: Bria",
        "task": "Remove Background",
        "model": "BRIA",
        "freshness": "established",
        "usage": 41,
        "recommend": "not_recommended",
        "description": "Upload an image to automatically remove its background using the BRIA API, which returns the result with a transparent background. This utility workflow processes the request asynchronously and is ideal for compositing, product photography, or any task needing clean subject isolation. Since it uses a third-party API, execution time depends on server-side response.",
        "io": {
          "inputs": [
            "image: Input image to remove background from"
          ],
          "outputs": [
            "image: Output image with transparent background"
          ]
        },
        "minComfyUIVersion": "0.14.0",
        "capabilities": {
          "workflow": [
            "api",
            "background-removal"
          ]
        }
      },
      {
        "name": "utility_bria_increase_resolution",
        "title": "Image Upscale: Bria Increase Resolution",
        "task": "Image Upscale",
        "model": "BRIA",
        "freshness": "recent",
        "usage": 65,
        "recommend": "low",
        "description": "Upscale any image to higher resolutions using Bria's dedicated resolution enhancement API, which preserves original content without adding new details or regenerating the image. The workflow accepts a single input image and produces one enhanced, higher-resolution output, supporting JPEG and PNG files in RGB, RGBA, or CMYK color modes up to a maximum area of 8192x8192 pixels. Built-in content moderation scans both the input image and generated output to ensure safe, policy-compliant processing. Ideal for preparing images for print, enlarging photos for digital display, or increasing resolution of assets for professional design work.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.34.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-upscale"
          ]
        }
      },
      {
        "name": "api_clarity_ai_crystal_upscale",
        "title": "Clarity AI: Crystal Upscaler",
        "task": "Image Upscale",
        "model": "Crystal Upscaler",
        "freshness": "new",
        "usage": 219,
        "recommend": "medium",
        "description": "Upscale and enhance images with the precision-focused Crystal model, preserving real detail while adding resolution. Takes one image in and returns one upscaled image plus a side-by-side comparison view. Ideal for portraits, product shots, and print-ready enlargements where faithful detail matters.",
        "io": {
          "inputs": [
            "image: Input image"
          ],
          "outputs": [
            "image: Upscaled image"
          ]
        },
        "minComfyUIVersion": "0.39.2",
        "capabilities": {
          "workflow": [
            "api",
            "image-upscale"
          ],
          "model_options": {
            "ClarityCrystalUpscaleNode": [
              "crystal-upscaler"
            ]
          }
        }
      },
      {
        "name": "api_magnific_image_upscale_precise",
        "title": "Image Upscale: Magnific Precise",
        "task": "Image Upscale",
        "model": "Magnific",
        "freshness": "established",
        "usage": 72,
        "recommend": "low",
        "description": "This workflow precisely upsamples an input image using the Magnific model, producing a high-resolution version with enhanced detail while preserving the original structure. It accepts a single image and returns the upscaled output. Execution relies on a third-party API call, so response time depends on server-side processing.",
        "io": {
          "inputs": [
            "image: Input image to upscale precisely"
          ],
          "outputs": [
            "image: Output precisely upscaled high-resolution image"
          ]
        },
        "minComfyUIVersion": "0.11.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-upscale"
          ]
        }
      },
      {
        "name": "api_bytedance_seedream_5_0_layer_separation",
        "title": "Layer Decomposition: Seedream 5.0 Pro",
        "task": "Image Generation",
        "model": "Seedream 5.0 Pro",
        "freshness": "recent",
        "usage": 29,
        "recommend": "not_recommended",
        "description": "Split any image into independent, editable layers — subject, product, background, and text — with a single text instruction using Seedream 5.0 Pro. A single poster can yield 10+ layers, with design-grade text separation supporting multilingual headlines and CJK fonts. Ideal for brand design, e-commerce asset creation, and layout iteration.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.31.0",
        "capabilities": {
          "workflow": [
            "api",
            "layer-decompose"
          ],
          "model_options": {
            "ByteDanceSeedreamLayerSeparationNodeV2": [
              "seedream 5.0 pro",
              "seedream 5.0 flash"
            ]
          }
        }
      },
      {
        "name": "utility_topaz_illustration_upscale",
        "title": "Image Upscale: Topaz Illustration",
        "task": "Image Upscale",
        "model": "Topaz",
        "freshness": "established",
        "usage": 29,
        "recommend": "not_recommended",
        "description": "Upload an illustration to upscale using Topaz. Generate a smooth, stylized image with accurate colors, though some artifacts may appear.",
        "io": {
          "inputs": [
            "image: Input image"
          ],
          "outputs": [
            "image: Upscaled image"
          ]
        },
        "minComfyUIVersion": "0.30.2",
        "capabilities": {
          "workflow": [
            "api",
            "image-upscale"
          ],
          "model_options": {
            "TopazImageEnhanceV2": [
              "Reimagine",
              "Bloom 2",
              "Wonder 3.5"
            ]
          }
        }
      },
      {
        "name": "api_topaz_image_enhance_bloom2",
        "title": "Image Upscale: Topaz Bloom 2",
        "task": "Image Upscale",
        "model": "Topaz",
        "freshness": "current",
        "usage": 63,
        "recommend": "low",
        "description": "Enhance AI-generated images with Bloom 2, a creative upscaling model that adds new detail, texture, and visual elements while offering a Creativity control from 1–9 for subtle refinement or bold reinterpretation. It accepts 1 input image and produces 1 enhanced output, with optional Face Recovery, Color Preservation, and grain settings for fine-tuned results. Ideal for improving AI artwork and illustrations, adding realism to CG renders, and refining facial detail in generated portraits.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.30.2",
        "capabilities": {
          "workflow": [
            "api",
            "image-upscale"
          ],
          "model_options": {
            "TopazImageEnhanceV2": [
              "Reimagine",
              "Bloom 2",
              "Wonder 3.5"
            ]
          }
        }
      },
      {
        "name": "api_wavespeed_image_upscale",
        "title": "Image Upscale: WaveSpeed",
        "task": "Image Upscale",
        "model": "WaveSpeed",
        "freshness": "established",
        "usage": 26,
        "recommend": "not_recommended",
        "description": "Upload an image to upscale its resolution using the WaveSpeed API. The workflow generates enhanced 2K, 4K, or 8K output that preserves details and reduces artifacts, supporting JPEG, PNG, or WEBP formats. Execution time depends on server-side response as this calls a third-party API.",
        "io": {
          "inputs": [
            "image: Input image to upscale with WaveSpeed"
          ],
          "outputs": [
            "image: Output upscaled image at 2K, 4K, or 8K resolution"
          ]
        },
        "minComfyUIVersion": "0.11.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-upscale"
          ],
          "model_options": {
            "WavespeedImageUpscaleNode": [
              "SeedVR2",
              "Ultimate"
            ]
          }
        }
      },
      {
        "name": "api_wavespeed_seedvr2_ai_image_fix",
        "title": "Image Upscale: WaveSpeed SeedVR2",
        "task": "Image Upscale",
        "model": "SeedVR2",
        "freshness": "established",
        "usage": 26,
        "recommend": "not_recommended",
        "description": "This workflow uses ByteDance's fast SeedVR2 video upscaler to restore and upscale AI-generated images up to 4K resolution while preserving colors and sharpness. It minimizes common artifacts and outputs a high-quality enlarged image. Execution relies on a third-party API, with time depending on server response.",
        "io": {
          "inputs": [
            "image: Input AI-generated image to restore and upscale with SeedVR2"
          ],
          "outputs": [
            "image: Output restored and upscaled image with minimal artifacts"
          ]
        },
        "minComfyUIVersion": "0.28.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-upscale"
          ],
          "model_options": {
            "WavespeedImageUpscaleNode": [
              "SeedVR2",
              "Ultimate"
            ]
          }
        }
      },
      {
        "name": "utility_nanobanana_pro_product_upscale",
        "title": "Image Upscale: Nano Banana Pro Product",
        "task": "Image Upscale",
        "model": "Nano Banana Pro",
        "freshness": "established",
        "usage": 55,
        "recommend": "low",
        "description": "Upload a product image to enhance detail and sharpness with the Nano Banana Pro model, known for complex image editing. The workflow generates an upscaled, high-quality output with improved typography and realistic coherence for e-commerce use. This workflow calls a third-party API, with execution time depending on server response.",
        "io": {
          "inputs": [
            "image: Input product image to enhance and upscale"
          ],
          "outputs": [
            "image: Output upscaled product image with improved detail and sharpness"
          ]
        },
        "minComfyUIVersion": "0.3.71",
        "capabilities": {
          "workflow": [
            "api",
            "image-upscale"
          ],
          "model_options": {
            "GeminiImage2Node": [
              "gemini-3-pro-image-preview",
              "Nano Banana 2 (Gemini 3.1 Flash Image)"
            ]
          }
        }
      },
      {
        "name": "utility_nanobanana_pro_illustration_upscale",
        "title": "Image Upscale: Nano Banana Pro Illustration",
        "task": "Image Upscale",
        "model": "Gemini3 Pro Image Preview",
        "freshness": "established",
        "usage": 25,
        "recommend": "not_recommended",
        "description": "Upload a stylized art or illustration to creatively upscale using the Nano Banana Pro model powered by Gemini 3 Pro. The workflow generates a 4K output with enhanced detail, ideal for marketing or gaming assets. This workflow calls a third-party API.",
        "io": {
          "inputs": [
            "image: Input stylized art or illustration to upscale creatively"
          ],
          "outputs": [
            "image: Output creatively upscaled 4K illustration with enhanced detail"
          ]
        },
        "minComfyUIVersion": "0.3.71",
        "capabilities": {
          "workflow": [
            "api",
            "image-upscale"
          ],
          "model_options": {
            "GeminiImage2Node": [
              "gemini-3-pro-image-preview",
              "Nano Banana 2 (Gemini 3.1 Flash Image)"
            ]
          }
        }
      },
      {
        "name": "image_marigold_v2_surface_normal_estimation",
        "title": "Surface Normal Estimation: Marigold V2",
        "task": "Image Generation",
        "model": "Marigold v2",
        "freshness": "recent",
        "usage": 33,
        "recommend": "not_recommended",
        "description": "Estimate surface normals from an input image using Marigold V2. The workflow generates a normal map representing the orientation of surfaces in the scene, useful for 3D reconstruction, relighting, and geometry-aware image processing.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.36.0",
        "capabilities": {
          "workflow": [
            "normal-estimation",
            "int8"
          ]
        }
      },
      {
        "name": "image_marigold_v2_albedo_estimation",
        "title": "Albedo Estimation: Marigold V2",
        "task": "Image Generation",
        "model": "Marigold v2",
        "freshness": "recent",
        "usage": 23,
        "recommend": "not_recommended",
        "description": "Estimate albedo from an input image using Marigold V2. The workflow separates surface color from lighting and shading, producing an albedo map useful for relighting, material analysis, and 3D asset creation.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.36.0",
        "capabilities": {
          "workflow": [
            "albedo-estimation",
            "int8"
          ]
        }
      },
      {
        "name": "utility_sdpose_ood_image_to_pose",
        "title": "Pose Map: SDPose-OOD",
        "task": "Image Generation",
        "model": "SDPose-OOD",
        "freshness": "established",
        "usage": 71,
        "recommend": "low",
        "description": "Upload an image to extract pose keypoints and generate a corresponding pose map using the SDPose-OOD model.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.16.0",
        "capabilities": {
          "workflow": [
            "pose-map"
          ]
        }
      },
      {
        "name": "api_bfl_flux1_expand_image",
        "title": "Image Outpainting: Flux.1",
        "task": "Image Generation",
        "model": "FLUX",
        "freshness": "established",
        "usage": 32,
        "recommend": "not_recommended",
        "description": "Use Flux.1 to expand any image beyond its original canvas via outpainting. This fast model generates extended content that matches the input style, though colors and textures are less realistic. The workflow calls a third-party API, so runtime depends on server response.",
        "io": {
          "inputs": [
            "image: Input image to expand via outpainting"
          ],
          "outputs": [
            "image: Output outpainted image with extended canvas content"
          ]
        },
        "minComfyUIVersion": "0.3.32",
        "capabilities": {
          "workflow": [
            "api",
            "outpainting"
          ]
        }
      },
      {
        "name": "utility_hitpaw_general_image_enhance",
        "title": "Image Upscale: HitPaw Portrait",
        "task": "Image Upscale",
        "model": "HitPaw",
        "freshness": "established",
        "usage": 37,
        "recommend": "not_recommended",
        "description": "Upload a low-quality portrait image and select an upscaling model via the HitPaw General Image Enhance API. The workflow generates a high-fidelity portrait with improved skin textures and facial features. This workflow calls a third-party API; execution time depends on server-side response.",
        "io": {
          "inputs": [
            "image: Input portrait image to enhance with HitPaw"
          ],
          "outputs": [
            "image: Output enhanced portrait with improved skin textures and facial features"
          ]
        },
        "minComfyUIVersion": "0.12.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-upscale"
          ],
          "model_options": {
            "HitPawGeneralImageEnhance": [
              "generative_portrait",
              "generative"
            ]
          }
        }
      },
      {
        "name": "api_beeble_switchx_image_edit",
        "title": "Relight: Beeble SwitchX",
        "task": "Image Edit",
        "model": "SwitchX",
        "freshness": "current",
        "usage": 14,
        "recommend": "not_recommended",
        "description": "Upload a source image and optional reference image for style/lighting along with a text prompt. Beeble SwitchX edits masked regions to match the prompt and reference while relighting preserved areas harmoniously. Outputs include the composited result, a before/after comparison, and an alpha matte cutout. This workflow calls a third‑party API, so response time depends on server availability.",
        "io": {
          "inputs": [
            "image×2: Source image to edit and optional reference image for style/lighting",
            "text: Text prompt describing desired image edits"
          ],
          "outputs": [
            "image: Edited image with prompt-guided regions and preserved source areas"
          ]
        },
        "minComfyUIVersion": "0.23.0",
        "capabilities": {
          "workflow": [
            "api",
            "relight",
            "image-edit"
          ]
        }
      },
      {
        "name": "api_topaz_image_enhance",
        "title": "Image Upscale: Topaz Reimagine",
        "task": "Image Upscale",
        "model": "Reimagine",
        "freshness": "established",
        "usage": 42,
        "recommend": "not_recommended",
        "description": "Professional image enhancement using Topaz's Reimagine model with face enhancement and detail restoration.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.71",
        "capabilities": {
          "workflow": [
            "api",
            "image-upscale"
          ],
          "model_options": {
            "TopazImageEnhanceV2": [
              "Reimagine",
              "Bloom 2",
              "Wonder 3.5"
            ]
          }
        }
      },
      {
        "name": "utility-topaz_landscape_upscaler",
        "title": "Image Upscale: Topaz Reimagine Landscape",
        "task": "Image Upscale",
        "model": "Topaz",
        "freshness": "established",
        "usage": 37,
        "recommend": "not_recommended",
        "description": "Upload a landscape image to upscale and enhance details. Generate a high-resolution output optimized for natural scenery using Topaz Reimagine.",
        "io": {
          "inputs": [
            "image: Input image"
          ],
          "outputs": [
            "image: Upscaled image"
          ]
        },
        "minComfyUIVersion": "0.3.71",
        "capabilities": {
          "workflow": [
            "api",
            "image-upscale"
          ],
          "model_options": {
            "TopazImageEnhanceV2": [
              "Reimagine",
              "Bloom 2",
              "Wonder 3.5"
            ]
          }
        }
      },
      {
        "name": "api_bria_eraser",
        "title": "Object Removal: Bria Eraser",
        "task": "Image Generation",
        "model": "BRIA",
        "freshness": "recent",
        "usage": 34,
        "recommend": "not_recommended",
        "description": "Remove unwanted objects or regions from images with Bria's Eraser, which fills the masked area with surrounding context while leaving the rest of the image pixel-perfect. Upload 1 image, paint over the target object with a brush mask, and receive 1 cleaned output alongside a before-and-after comparison. Ideal for cleaning up photobombs, removing watermarks or text, and decluttering product or real-estate photos.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.33.0",
        "capabilities": {
          "workflow": [
            "api",
            "inpainting"
          ]
        }
      },
      {
        "name": "utility_interpolation_image_upscale",
        "title": "Image Upscale: Traditional Interpolation",
        "task": "Image Upscale",
        "model": "None",
        "freshness": "established",
        "usage": 18,
        "recommend": "not_recommended",
        "description": "This workflow upscales images using traditional interpolation algorithms for faster processing with minimal detail changes. Simply upload an image and select the interpolation method to produce a higher‑resolution output. No AI model is involved, making it ideal for quick, deterministic upscaling tasks.",
        "io": {
          "inputs": [
            "image: Image to upscale"
          ],
          "outputs": [
            "image: Upscaled high-resolution image"
          ]
        },
        "minComfyUIVersion": "0.9.0",
        "capabilities": {
          "workflow": [
            "image-upscale"
          ]
        }
      },
      {
        "name": "utility_recraft_creative_image_upscale",
        "title": "Image Upscale: Recraft Creative",
        "task": "Image Upscale",
        "model": "Recraft",
        "freshness": "established",
        "usage": 13,
        "recommend": "not_recommended",
        "description": "Upload an image to creatively upscale using the Recraft API. The AI reconstructs details and improves textures for a higher-quality output. Input limited to 5MB and 2048x2048 resolution. This workflow calls a third-party API, so execution time depends on server response.",
        "io": {
          "inputs": [
            "image: Input image to creatively upscale (max 5MB, 2048x2048)"
          ],
          "outputs": [
            "image: Output creatively upscaled image with reconstructed details and improved textures"
          ]
        },
        "minComfyUIVersion": "0.9.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-upscale"
          ]
        }
      },
      {
        "name": "utility_recraft_crisp_image_upscale",
        "title": "Image Upscale: Recraft Crisp",
        "task": "Image Upscale",
        "model": "Recraft",
        "freshness": "established",
        "usage": 23,
        "recommend": "not_recommended",
        "description": "Upload a low-resolution or noisy image to the Recraft Crisp Upscale API for restoration and upscaling. It produces a high-resolution output with enhanced details, reduced artifacts, and a natural look. Input must be under 5MB and 2048×2048; execution time depends on the API response.",
        "io": {
          "inputs": [
            "image: Input low-resolution or noisy image to restore and upscale (max 5MB, 2048x2048)"
          ],
          "outputs": [
            "image: Output restored high-resolution image with reduced artifacts and enhanced details"
          ]
        },
        "minComfyUIVersion": "0.9.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-upscale"
          ]
        }
      },
      {
        "name": "image_qwen_image_layered_control",
        "title": "Layer Decomposition: Qwen-Image-Layered Control",
        "task": "Image Generation",
        "model": "Qwen-Image-Layered",
        "freshness": "established",
        "usage": 27,
        "recommend": "not_recommended",
        "description": "Describe the elements you want in the image and decompose it precisely from the image.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.6.0",
        "capabilities": {
          "workflow": [
            "layer-decompose"
          ]
        }
      },
      {
        "name": "api_magnific_image_relight",
        "title": "Relight: Magnific",
        "task": "Relight",
        "model": "Magnific",
        "freshness": "established",
        "usage": 32,
        "recommend": "not_recommended",
        "description": "Upload a source image and a reference lighting image. The Magnific model generates a relit version of the source, applying the lighting, shadows, and ambiance from the reference. This workflow calls a third-party API; execution time depends on server response.",
        "io": {
          "inputs": [
            "image×2: Source image to relight and reference lighting image to apply"
          ],
          "outputs": [
            "image: Output relit image with lighting from the reference image"
          ]
        },
        "minComfyUIVersion": "0.11.0",
        "capabilities": {
          "workflow": [
            "api",
            "relight"
          ]
        }
      },
      {
        "name": "image_lotus_depth_v1_1",
        "title": "Depth Estimation: Lotus",
        "task": "Text to Image",
        "model": "SD1.5",
        "freshness": "established",
        "usage": 10,
        "recommend": "not_recommended",
        "description": "Run Lotus Depth in ComfyUI for zero-shot, efficient monocular depth estimation with high detail retention.",
        "io": {
          "inputs": [
            "image: Input image"
          ],
          "outputs": [
            "image: Depth map"
          ]
        },
        "minComfyUIVersion": "0.3.42",
        "capabilities": {
          "workflow": [
            "text-to-image"
          ]
        }
      },
      {
        "name": "utility_moge_depth_estimation",
        "title": "Depth Estimation: MoGe",
        "task": "Depth Estimation",
        "model": "MoGe",
        "freshness": "current",
        "usage": 13,
        "recommend": "not_recommended",
        "description": "Upload a single RGB image and adjust inference resolution and batch size. Generate a colored depth preview and raw depth map,",
        "io": {
          "inputs": [
            "image: Input image"
          ],
          "outputs": [
            "image: Depth map"
          ]
        },
        "minComfyUIVersion": "0.22.0",
        "capabilities": {
          "workflow": [
            "controlnet"
          ]
        }
      },
      {
        "name": "utility_sdpose_multi_person",
        "title": "Pose Detection: SDPose Multi-Person",
        "task": "Image Generation",
        "model": "SDPose-OOD",
        "freshness": "established",
        "usage": 26,
        "recommend": "not_recommended",
        "description": "Upload an image to detect human poses. Supports detection for both single individuals and multiple people within the same scene.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.19.0",
        "capabilities": {
          "workflow": [
            "pose-map"
          ]
        }
      },
      {
        "name": "api_flux_erase_image",
        "title": "Object Removal: Flux",
        "task": "Image Generation",
        "model": "FLUX",
        "freshness": "current",
        "usage": 18,
        "recommend": "not_recommended",
        "description": "Upload an image with a masked area, and Flux erases the object while reconstructing the background without a text prompt. This fast, inpainting-focused workflow executes via a third-party API, so runtime depends on server response.",
        "io": {
          "inputs": [
            "image: Input image with masked area to erase"
          ],
          "outputs": [
            "image: Output image with masked area removed and background reconstructed"
          ]
        },
        "minComfyUIVersion": "0.23.0",
        "capabilities": {
          "workflow": [
            "api",
            "inpainting"
          ]
        }
      }
    ]
  },
  {
    "category": "Video",
    "description": "General-purpose workflow templates for native video generation, including text-to-video, image-to-video, and other core video workflows supported out of the box.",
    "templates": [
      {
        "name": "video_minimax_h3_i2v",
        "title": "MiniMax H3: Image to Video",
        "task": "Image to Video",
        "model": "MiniMax H3",
        "freshness": "current",
        "usage": 19644,
        "recommend": "highly_recommended",
        "description": "Turn a starting image into a short, high-quality video clip with MiniMax H3, a unified multimodal model that generates visuals and synchronized audio in a single pass. The workflow takes 1 input image and produces 1 video output, with optional text or reference media to steer motion, style, or sound. Ideal for rapid concept visualization, character-driven animation tests, and creating shareable social media clips with built-in sound design.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.30.0",
        "capabilities": {
          "workflow": [
            "image-to-video",
            "int8"
          ]
        }
      },
      {
        "name": "video_minimax_h3_r2v",
        "title": "MiniMax H3: Reference to Video",
        "task": "Reference to Video",
        "model": "MiniMax H3",
        "freshness": "current",
        "usage": 6777,
        "recommend": "highly_recommended",
        "description": "Generate a new video by feeding MiniMax H3 a reference image, up to nine images, three videos, and three audio clips alongside a text prompt, letting the model match character identity, motion, camera, voice, or style from those sources while producing native stereo audio in a single pass. Output is a single MP4 video at up to 2K resolution and 24fps for roughly 15 seconds. Ideal for character-consistent animation, style-transfer video generation, and audio-synced content creation.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.30.0",
        "capabilities": {
          "workflow": [
            "reference-to-video",
            "int8"
          ]
        }
      },
      {
        "name": "video_minimax_h3_t2v",
        "title": "MiniMax H3: Text to Video",
        "task": "Text to Video",
        "model": "MiniMax H3",
        "freshness": "current",
        "usage": 4356,
        "recommend": "highly_recommended",
        "description": "Generate video with native stereo audio directly from a text prompt using MiniMax H3, an omni-modal model that synthesizes voice, sound effects, and music in a single pass. Output is up to 2K resolution, 24fps, and roughly 15 seconds long, with no file inputs required — just a prompt to one video output. Ideal for rapid concept visualization, social media content creation, and audio-synced storytelling.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.30.0",
        "capabilities": {
          "workflow": [
            "text-to-video",
            "int8"
          ]
        }
      },
      {
        "name": "api_google_gemini_omni_flash_1_2_t2v",
        "title": "Gemini Omni 1.2 Flash: Text to Video",
        "task": "Text to Video",
        "model": "Gemini Omni 1.2 Flash",
        "freshness": "new",
        "usage": 0,
        "recommend": "medium",
        "description": "Write the shot and get a 5-second 16:9 clip with synchronized audio: ambience, footsteps and UI clicks come out of the same pass. Gemini Omni 1.2 Flash keeps a fixed cast and layout across the clip, so game-lobby key art, product beats and menu shots hold together.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.44.1",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-video"
          ],
          "model_options": {
            "GeminiVideoOmniV2": [
              "Omni Flash 1.1",
              "Omni Flash"
            ]
          }
        }
      },
      {
        "name": "api_google_gemini_omni_flash_1_2_i2v",
        "title": "Gemini Omni 1.2 Flash: Image to Video",
        "task": "Image to Video",
        "model": "Gemini Omni 1.2 Flash",
        "freshness": "new",
        "usage": 0,
        "recommend": "medium",
        "description": "Animate a still: the image becomes the first frame and the prompt carries the motion, camera and sound. Gemini Omni Flash renders picture and audio in one pass, so ambience and voices arrive with the clip.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.44.1",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video"
          ],
          "model_options": {
            "GeminiVideoOmniV2": [
              "Omni Flash 1.1",
              "Omni Flash"
            ]
          }
        }
      },
      {
        "name": "api_google_gemini_omni_flash_1_2_r2v",
        "title": "Gemini Omni 1.2 Flash: Reference to Video",
        "task": "Reference to Video",
        "model": "Gemini Omni 1.2 Flash",
        "freshness": "new",
        "usage": 0,
        "recommend": "medium",
        "description": "Split a character across two references: image 1 carries the silhouette and armour, image 2 locks the face and detail, and the same identity holds through the whole clip. Gemini Omni Flash renders picture and sound in one pass, so the scene and its audio arrive together.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.44.1",
        "capabilities": {
          "workflow": [
            "api",
            "reference-to-video"
          ],
          "model_options": {
            "GeminiVideoOmniV2": [
              "Omni Flash 1.1",
              "Omni Flash"
            ]
          }
        }
      },
      {
        "name": "video_pika_h3_flex_t2v",
        "title": "Pika H3 Flex: Text to Video",
        "task": "Text to Video",
        "model": "MiniMax H3",
        "freshness": "new",
        "usage": 873,
        "recommend": "high",
        "description": "Generate a video with synchronized audio from a text prompt with Pika H3 Flex, a Pika post-trained adapter for MiniMax H3 that lands a finished 1344x768, 124-frame clip in six sampling steps. The adapter runs at strength 1.0 on the int8 MiniMax H3 base with the lcm sampler and Pika's own six-value sigma list, so a full render takes about a minute instead of a 20-step run. Ideal for fast prompt exploration and concept iteration on H3, with no image input to prepare.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.44.0",
        "capabilities": {
          "workflow": [
            "text-to-video"
          ]
        }
      },
      {
        "name": "video_pika_h3_flex_fl2va",
        "title": "Pika H3 Flex: FL2VA",
        "task": "FLF2V",
        "model": "MiniMax H3",
        "freshness": "new",
        "usage": 411,
        "recommend": "medium",
        "description": "Animate between two still frames with Pika H3 Flex: the first image opens the shot and the second closes it, while the prompt describes the path between them. The two frames are locked to the start and the end of the clip, the adapter runs at strength 1.0 on the int8 MiniMax H3 FL2VA base with the lcm sampler and Pika's six-step sigma list, and the output size is taken from the first frame and snapped to a multiple of 32 automatically, so portrait sources stay portrait.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.44.0",
        "capabilities": {
          "workflow": [
            "flf2v",
            "image-to-video"
          ]
        }
      },
      {
        "name": "video_pika_h3_flex_ref2va",
        "title": "Pika H3 Flex: Reference to Video",
        "task": "Reference to Video",
        "model": "MiniMax H3",
        "freshness": "new",
        "usage": 1822,
        "recommend": "top",
        "description": "Generate a video with synchronized audio from reference images with Pika H3 Flex, keeping the identity, costume and look of the references while the prompt carries the story. Two image inputs are wired as Picture 1 and Picture 2, and the Pika adapter runs at strength 1.0 with the lcm sampler and a six-step sigma list on the MiniMax H3 Ref2VA base. Ideal for character-consistent short clips with native sound.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.44.0",
        "capabilities": {
          "workflow": [
            "reference-to-video",
            "lora"
          ]
        }
      },
      {
        "name": "video_fasth3_ominiref_r2v",
        "title": "FastH3 OmniRef: Reference to Video",
        "task": "Reference to Video",
        "model": "MiniMax H3",
        "freshness": "new",
        "usage": 2141,
        "recommend": "top",
        "description": "Generate synchronized video with matching audio from a text prompt plus a reference image using the FastH3 OmniRef eight-step distilled model, which renders both streams in just eight sampling steps. Takes 1 input image and produces 1 video output. Ideal for animating reference characters, creating talking avatars, and rapid social media video prototyping.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.43.0",
        "capabilities": {
          "workflow": [
            "reference-to-video"
          ]
        }
      },
      {
        "name": "video_ltx2_3_i2v",
        "title": "LTX-2.3: Image to Video",
        "task": "Image to Video",
        "model": "LTX-2.3",
        "freshness": "established",
        "usage": 18605,
        "recommend": "highly_recommended",
        "description": "Upload an image to generate a video with improved motion consistency and fine details, including enhanced support for portrait formats and audio quality.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.16.4",
        "capabilities": {
          "workflow": [
            "image-to-video"
          ]
        }
      },
      {
        "name": "video_wan2_2_14B_i2v",
        "title": "Wan 2.2 14B Image to Video",
        "task": "Image to Video",
        "model": "Wan",
        "freshness": "established",
        "usage": 9549,
        "recommend": "highly_recommended",
        "description": "Transform static images into dynamic videos with precise motion control and style preservation using Wan 2.2.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.16.4",
        "capabilities": {
          "workflow": [
            "image-to-video"
          ]
        }
      },
      {
        "name": "video_ltx2_5_i2v",
        "title": "LTX-2.5: Image to Video",
        "task": "Image to Video",
        "model": "LTX-2.5",
        "freshness": "recent",
        "usage": 4490,
        "recommend": "highly_recommended",
        "description": "Generate video from a single image using LTX-2.5, producing a high-fidelity clip with industry-leading pixel quality and native multishot capability that holds character, environment, and lighting across connected scenes. Ideal for cinematic production, VFX shot creation, and content pipelines requiring fast, locally deployable video generation.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.32.0",
        "capabilities": {
          "workflow": [
            "image-to-video",
            "int8"
          ]
        }
      },
      {
        "name": "video_ltx2_5_t2v",
        "title": "LTX-2.5: Text to Video",
        "task": "Text to Video",
        "model": "LTX-2.5",
        "freshness": "recent",
        "usage": 1545,
        "recommend": "top",
        "description": "Generate cinematic video from a text prompt using LTX-2.5, a production-ready diffusion model that outputs a single video file with optional spatial upscaling for enhanced resolution. Ideal for film previsualization, commercial storyboards, and high-volume content creation pipelines.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.32.0",
        "capabilities": {
          "workflow": [
            "text-to-video",
            "int8"
          ]
        }
      },
      {
        "name": "video_fastvideo_fasth3_i2v",
        "title": "FastVideo FastH3: Image to Video",
        "task": "Image to Video",
        "model": "MiniMax H3",
        "freshness": "recent",
        "usage": 1657,
        "recommend": "top",
        "description": "Generate video with synchronized native audio from a single still image using MiniMax H3 and the FastH3 8-step checkpoint. One image produces one video output with matching voice, sound effects, and music in a single pass. Ideal for animating portraits, bringing product shots to life, and quick social media clips.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.36.0",
        "capabilities": {
          "workflow": [
            "image-to-video",
            "int8"
          ]
        }
      },
      {
        "name": "video_fastvideo_fasth3_t2v",
        "title": "FastVideo FastH3: Text to Video",
        "task": "Text to Video",
        "model": "MiniMax H3",
        "freshness": "recent",
        "usage": 464,
        "recommend": "medium",
        "description": "Generate synchronized video and audio directly from text using a distilled eight-step model, producing one video output with no file inputs required. Ideal for rapid text-to-video prototyping, audio-visual content creation, and creative storytelling where fast generation matters more than maximum detail.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.36.0",
        "capabilities": {
          "workflow": [
            "text-to-video",
            "int8"
          ]
        }
      },
      {
        "name": "api_seedance2_5_i2v_1080p",
        "title": "Seedance2.5 1080P: Image to Video",
        "task": "Image to Video",
        "model": "Seedance 2.5",
        "freshness": "recent",
        "usage": 1713,
        "recommend": "top",
        "description": "Generate cinematic 1080P video clips from a single image using Seedance 2.5, animating the still into a moving scene with synchronized sound, music, and dialogue for clips up to 30 seconds across any aspect ratio. Ideal for breathing life into product shots, extending concept art into motion, and creating engaging social media content from photographs.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.33.2",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video"
          ],
          "model_options": {
            "ByteDance2ReferenceNodeV2": [
              "Seedance 2.5",
              "Seedance 2.5 Draft",
              "Seedance 2.0",
              "Seedance 2.0 Fast",
              "Seedance 2.0 Mini"
            ]
          }
        }
      },
      {
        "name": "api_seedance2_5_r2v",
        "title": "Seedance 2.5: Reference to Video",
        "task": "Reference to Video",
        "model": "Seedance 2.5",
        "freshness": "recent",
        "usage": 3038,
        "recommend": "highly_recommended",
        "description": "Generate reference-driven videos with Seedance 2.5 by supplying up to 20 images, 6 video clips, and 6 audio clips to control character consistency, visual style, and motion rhythm in a single 4–30 second output with synchronized sound and lip-synced dialogue. Ideal for e-commerce product showcases, brand advertising, and character-driven short series.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.33.2",
        "capabilities": {
          "workflow": [
            "api",
            "reference-to-video"
          ],
          "model_options": {
            "ByteDance2ReferenceNodeV2": [
              "Seedance 2.5",
              "Seedance 2.5 Draft",
              "Seedance 2.0",
              "Seedance 2.0 Fast",
              "Seedance 2.0 Mini"
            ]
          }
        }
      },
      {
        "name": "video_ltx2_3_ia2v",
        "title": "LTX-2.3: Image Audio to Video",
        "task": "Image to Video",
        "model": "LTX-2.3",
        "freshness": "established",
        "usage": 1274,
        "recommend": "top",
        "description": "Upload an image and an audio file to generate a high-quality video with synchronized lip movements using LTX-2.3.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.16.4",
        "capabilities": {
          "workflow": [
            "image-to-video",
            "lip-sync"
          ]
        }
      },
      {
        "name": "api_flux3_max_i2v",
        "title": "FLUX 3 Max: Image to Video",
        "task": "Image to Video",
        "model": "FLUX 3 Video",
        "freshness": "new",
        "usage": 527,
        "recommend": "high",
        "description": "Animate a single still image into a short video clip using FLUX 3 Max, taking one input image and producing one video output. Ideal for bringing product shots, portraits, and concept art to life, generating social media motion content, and rapid prototyping of animated scenes.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.42.2",
        "capabilities": {
          "workflow": [
            "image-to-video",
            "api"
          ]
        }
      },
      {
        "name": "api_flux3_max_t2v",
        "title": "FLUX 3 Max: Text to Video",
        "task": "Text to Video",
        "model": "FLUX 3 Video",
        "freshness": "new",
        "usage": 114,
        "recommend": "medium",
        "description": "Generate a video with synchronized audio from a single text prompt with FLUX 3 Max, which reasons about and expands the prompt before rendering. Outputs one clip of 5 to 20 seconds at 720p or 1080p, at any aspect ratio from 9:16 to 21:9, with ambient sound, speech, and effects created alongside the picture and no input files required. Ideal for cinematic concept clips, ad mockups, and social videos that need matching sound design.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.42.2",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-video"
          ]
        }
      },
      {
        "name": "video_ltx2_3_t2v",
        "title": "LTX-2.3: Text to Video",
        "task": "Text to Video",
        "model": "LTX-2.3",
        "freshness": "established",
        "usage": 697,
        "recommend": "high",
        "description": "Generate a video from a text prompt, optionally using an image for reference. Receive a high-quality video with improved motion, audio, and detail, optimized for portrait or landscape formats.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.16.4",
        "capabilities": {
          "workflow": [
            "text-to-video"
          ]
        }
      },
      {
        "name": "video_ltx2_5_flf2v",
        "title": "LTX-2.5: FLF2V",
        "task": "FLF2V",
        "model": "LTX-2.5",
        "freshness": "recent",
        "usage": 939,
        "recommend": "high",
        "description": "Generate a video from first and last frame reference images using LTX-2.5, blending both frames into a single animated sequence. Ideal for creating transition shots, morphing between two subjects or scenes, and building multi-shot narratives from still references.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.32.0",
        "capabilities": {
          "workflow": [
            "flf2v",
            "int8"
          ]
        }
      },
      {
        "name": "api_seedance2_5_video_editing",
        "title": "Seedance2.5: Video Editing",
        "task": "Video Edit",
        "model": "Seedance 2.5",
        "freshness": "recent",
        "usage": 543,
        "recommend": "high",
        "description": "Edit existing video clips with Seedance 2.5 by describing changes in natural language, enabling local object removal, background swaps, scene replacements, and style transfers while preserving the original composition and motion. Feed in 1 video and receive 1 edited video output with temporally consistent results across frames. Ideal for e-commerce background replacement, stylized remakes of footage, and ad re-editing with targeted retouching.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.33.2",
        "capabilities": {
          "workflow": [
            "api",
            "video-edit"
          ],
          "model_options": {
            "ByteDance2ReferenceNodeV2": [
              "Seedance 2.5",
              "Seedance 2.5 Draft",
              "Seedance 2.0",
              "Seedance 2.0 Fast",
              "Seedance 2.0 Mini"
            ]
          }
        }
      },
      {
        "name": "api_seedance2_5_t2v",
        "title": "Seedance 2.5: Text to Video",
        "task": "Text to Video",
        "model": "Seedance 2.5",
        "freshness": "recent",
        "usage": 259,
        "recommend": "medium",
        "description": "Generate 30-second cinematic videos from text prompts with Seedance 2.5, producing synchronized sound and music alongside the footage. Input a single text prompt to create one video with precise camera control and multi-shot storytelling. Ideal for brand films, episodic series, and e-commerce content where quality matters more than budget.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.31.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-video"
          ],
          "model_options": {
            "ByteDance2TextToVideoNode": [
              "Seedance 2.5",
              "Seedance 2.5 Draft",
              "Seedance 2.0",
              "Seedance 2.0 Fast",
              "Seedance 2.0 Mini"
            ]
          }
        }
      },
      {
        "name": "api_seedance2_5_draft_r2v",
        "title": "Seedance2.5 Draft: Reference to Video",
        "task": "Reference to Video",
        "model": "Seedance 2.5",
        "freshness": "recent",
        "usage": 1069,
        "recommend": "top",
        "description": "Generate draft preview videos from reference images using Seedance 2.5's two-phase draft mode: stage one turns your references and prompt into a fast, low-cost 480p preview plus a reusable draft task ID, and stage two renders the 1080p final from that ID. Ideal for locking in a character or storyboard reference before committing to a full-quality render.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.37.3",
        "capabilities": {
          "workflow": [
            "api",
            "reference-to-video"
          ],
          "model_options": {
            "ByteDance2ReferenceNodeV2": [
              "Seedance 2.5",
              "Seedance 2.5 Draft",
              "Seedance 2.0",
              "Seedance 2.0 Fast",
              "Seedance 2.0 Mini"
            ]
          }
        }
      },
      {
        "name": "api_seedance2_5_draft_i2v",
        "title": "Seedance2.5 Draft: Image to Video",
        "task": "Image to Video",
        "model": "Seedance 2.5",
        "freshness": "recent",
        "usage": 179,
        "recommend": "low",
        "description": "Generate draft preview videos from a first frame image using Seedance 2.5's two-phase draft mode: stage one turns your image into a fast, low-cost 480p preview plus a reusable draft task ID, and stage two renders the 1080p final from that ID. Ideal for testing motion, camera and scene ideas on a still before committing to a full-quality render.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.37.3",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video"
          ],
          "model_options": {
            "ByteDance2FirstLastFrameNode": [
              "Seedance 2.5",
              "Seedance 2.5 Draft",
              "Seedance 2.0",
              "Seedance 2.0 Fast",
              "Seedance 2.0 Mini"
            ]
          }
        }
      },
      {
        "name": "video_wan_animate2",
        "title": "Wan Animate 2: Motion Transfer",
        "task": "Motion Control",
        "model": "Wan Animate",
        "freshness": "recent",
        "usage": 1757,
        "recommend": "top",
        "description": "Animate a still character using a driving video with Wan Animate 2, transferring motion directly from video frames without any pose extraction or skeleton preprocessing. The output preserves the reference character's identity while generating a fresh background and camera movement from your text prompt. Ideal for motion transfer, action mimicry, and character animation replacement projects.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.36.0",
        "capabilities": {
          "workflow": [
            "motion-control",
            "int8"
          ]
        }
      },
      {
        "name": "video_wan_animate2_distilled",
        "title": "Wan Animate 2 Distilled: Motion Transfer",
        "task": "Motion Control",
        "model": "Wan Animate",
        "freshness": "recent",
        "usage": 191,
        "recommend": "low",
        "description": "Animate a still character using a driving video with Wan Animate 2, transferring motion directly from video frames without any pose extraction or skeleton preprocessing. The output preserves the reference character's identity while generating a fresh background and camera movement from your text prompt. This distilled version delivers efficient, high-fidelity motion generation with strong identity preservation, ideal for motion transfer, action mimicry, and character animation replacement projects.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.31.0",
        "capabilities": {
          "workflow": [
            "motion-control",
            "int8"
          ]
        }
      },
      {
        "name": "api_higgsfield_genjutsu_motion_transfer",
        "title": "Higgsfield Genjutsu: Motion Transfer",
        "task": "Motion Control",
        "model": "Genjutsu 1.0",
        "freshness": "new",
        "usage": 279,
        "recommend": "medium",
        "description": "Recast an existing video with Higgsfield Genjutsu: the source clip keeps its motion, camera movement, timing and audio, while the cast, location and style are rebuilt from 1 to 8 reference images. Describe which reference replaces what in the optional prompt. Output length follows the source clip. Ideal for restyling footage, swapping the cast of a shot, and building scenes a shoot never covered.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.39.2",
        "capabilities": {
          "workflow": [
            "api",
            "character-replacement",
            "motion-control",
            "video-edit"
          ],
          "model_options": {
            "HiggsfieldMotionTransferNode": [
              "genjutsu-v1.0"
            ]
          }
        }
      },
      {
        "name": "video_wan21_scail2_character_replacement",
        "title": "SCAIL-2: Character Replacement",
        "task": "Video Generation",
        "model": "SCAIL-2",
        "freshness": "current",
        "usage": 882,
        "recommend": "high",
        "description": "Upload 1 reference character image and 1 driving video. Generate an animated video of the character matching the driving video's motion. Ideal for character animation, in-video character replacement, and motion transfer projects.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.25.0",
        "capabilities": {
          "workflow": [
            "character-replacement"
          ]
        }
      },
      {
        "name": "video_wan21_scail2_character_replacement_int8",
        "title": "SCAIL-2 Int8: Character Replacement",
        "task": "Video Generation",
        "model": "SCAIL-2",
        "freshness": "current",
        "usage": 615,
        "recommend": "high",
        "description": "Animate a reference character using a driving video or replace characters entirely with SCAIL-2 Int8, which delivers higher quality than fp8 while running faster. The workflow accepts one reference image and one driving video, outputting an animated character sequence with replaced identity. Ideal for cross-identity character swaps, animal-driven animation, and multi-character scenes without intermediate pose representations.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.25.0",
        "capabilities": {
          "workflow": [
            "character-replacement",
            "int8"
          ]
        }
      },
      {
        "name": "api_seedance2_0_r2v",
        "title": "Seedance 2.0: Reference to Video",
        "task": "Image to Video",
        "model": "Seedance 2.0",
        "freshness": "current",
        "usage": 537,
        "recommend": "high",
        "description": "Generate cinematic videos from reference images and text prompts. Preserve subject identity and composition while adding expressive motion with synchronized audio. Control camera movement and lighting through detailed prompts.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.26.2",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video"
          ],
          "model_options": {
            "ByteDance2ReferenceNodeV2": [
              "Seedance 2.5",
              "Seedance 2.5 Draft",
              "Seedance 2.0",
              "Seedance 2.0 Fast",
              "Seedance 2.0 Mini"
            ]
          }
        }
      },
      {
        "name": "api_vidu_q4_preview_r2v",
        "title": "Vidu Q4 Preview: Reference to Video",
        "task": "Reference to Video",
        "model": "Vidu Q4",
        "freshness": "new",
        "usage": 75,
        "recommend": "medium",
        "description": "Generate expressive videos from two reference images using Vidu Q4 Preview, a flagship model built for vivid character performance, dynamic camera language, and striking visual effects. Ideal for character-driven storytelling, cinematic action sequences, and effect-heavy scenes with explosions, smoke, or particles.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.39.2",
        "capabilities": {
          "workflow": [
            "api",
            "reference-to-video"
          ],
          "model_options": {
            "Vidu4ReferenceVideoNode": [
              "Vidu Q4 Preview"
            ]
          }
        }
      },
      {
        "name": "api_vidu_q4_preview_i2v",
        "title": "Vidu Q4 Preview: Image to Video",
        "task": "Image to Video",
        "model": "Vidu Q4",
        "freshness": "new",
        "usage": 42,
        "recommend": "medium",
        "description": "Generate cinematic videos from a single still image using Vidu Q4 Preview, an expressive flagship model built for lifelike character acting, dramatic camera movement, and striking visual effects. Submit 1 image and receive 1 finished video, with support for up to 15 reference images, 3 reference audio clips, and 2K or 4K output. Ideal for short films, marketing content, and social media storytelling.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.39.2",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video"
          ],
          "model_options": {
            "Vidu4ImageToVideoNode": [
              "Vidu Q4 Preview"
            ]
          }
        }
      },
      {
        "name": "api_seedance2_0_flf2v",
        "title": "Seedance2.0: First-Last-Frame to Video",
        "task": "FLF2V",
        "model": "Seedance 2.0",
        "freshness": "established",
        "usage": 263,
        "recommend": "medium",
        "description": "Provide the first and last frames of a video, and Seedance 2.0 generates a cinematic sequence interpolating between them. This workflow uses a third-party API and offers a choice between the standard or Fast variant for faster inference. The model produces smooth, film-like motion from the two keyframes.",
        "io": {
          "inputs": [
            "image×2: First and last frames to define the video sequence"
          ],
          "outputs": [
            "video: Generated video interpolating between the first and last frames"
          ]
        },
        "minComfyUIVersion": "0.26.2",
        "capabilities": {
          "workflow": [
            "api",
            "flf2v"
          ],
          "model_options": {
            "ByteDance2FirstLastFrameNode": [
              "Seedance 2.5",
              "Seedance 2.5 Draft",
              "Seedance 2.0",
              "Seedance 2.0 Fast",
              "Seedance 2.0 Mini"
            ]
          }
        }
      },
      {
        "name": "api_heygen_video_1_i2v",
        "title": "HeyGen Video 1.0: Image to Video",
        "task": "Image to Video",
        "model": "HeyGen",
        "freshness": "recent",
        "usage": 558,
        "recommend": "high",
        "description": "Generate a 5-15 second clip with synchronized dialogue and sound from a single image using HeyGen Video 1.0, which starts on that exact frame and keeps the image's aspect ratio, at 480p or 768p. The prompt drives both the action and the spoken line, and one video is output. Ideal for giving a portrait or product photo a voice, producing talking-head clips without a shoot, and bringing stills to life.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.38.1",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video"
          ],
          "model_options": {
            "HeyGenImageToVideoNode": [
              "heygen-video-1"
            ]
          }
        }
      },
      {
        "name": "api_wan3_0_i2v",
        "title": "Wan3.0: Image to Video",
        "task": "Image to Video",
        "model": "Wan3.0",
        "freshness": "recent",
        "usage": 475,
        "recommend": "medium",
        "description": "Create up to 30 seconds of high-fidelity video from a single image using Wan 3.0, which supports text, image, audio, and document inputs for versatile scene generation. Produces authentic, realistic footage with unique facial details and believable motion in one pass. Ideal for cinematic storytelling, product visualization, and documentary-style scene recreation.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.33.4",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video"
          ],
          "model_options": {
            "Wan3ImageToVideoApi": [
              "wan3.0-video",
              "wan3.0-video-prime"
            ]
          }
        }
      },
      {
        "name": "api_wan3_0_r2v",
        "title": "Wan3.0: Reference to Video",
        "task": "Reference to Video",
        "model": "Wan3.0",
        "freshness": "recent",
        "usage": 529,
        "recommend": "high",
        "description": "Generate a video from two reference images using Wan3.0's reference-to-video capability, which produces up to 30 seconds of footage with character consistency and lifelike visuals. Ideal for creating consistent character animations, scene replication from visual references, and producing audiovisual content from static imagery.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.33.4",
        "capabilities": {
          "workflow": [
            "api",
            "reference-to-video"
          ],
          "model_options": {
            "Wan3ReferenceToVideoApi": [
              "wan3.0-video",
              "wan3.0-video-prime"
            ]
          }
        }
      },
      {
        "name": "api_grok_video",
        "title": "Grok: Video generation",
        "task": "Image to Video",
        "model": "Grok",
        "freshness": "established",
        "usage": 260,
        "recommend": "medium",
        "description": "Use Grok via API to generate up to 15‑second videos with synchronized audio. Provide a text prompt or initial image; the workflow handles both text‑to‑video and image‑to‑video tasks. Fast, high‑quality output is delivered through the Grok model’s API capabilities.",
        "io": {
          "inputs": [
            "image: Optional initial image frame to animate",
            "text: Text prompt describing the video content"
          ],
          "outputs": [
            "video: Generated video with synchronized audio"
          ]
        },
        "minComfyUIVersion": "0.11.1",
        "capabilities": {
          "workflow": [
            "image-to-video",
            "text-to-video",
            "api"
          ],
          "model_options": {
            "GrokVideoNode": [
              "grok-imagine-video",
              "grok-imagine-video-1.5",
              "grok-imagine-video-1.5-lite"
            ]
          }
        }
      },
      {
        "name": "api_grok_imagine_video_1_5",
        "title": "Grok Imagine Video 1.5: Image to Video",
        "task": "Image to Video",
        "model": "Grok",
        "freshness": "current",
        "usage": 136,
        "recommend": "low",
        "description": "This workflow uses xAI’s Grok Imagine Video 1.5 API to animate a static image into a video, optionally guided by a text prompt. It produces native audio, realistic motion, and outputs up to 720p resolution. Execution depends on the third-party API, so processing time may vary.",
        "io": {
          "inputs": [
            "image: Input image to animate into a video",
            "text: Optional text prompt describing the desired motion or scene"
          ],
          "outputs": [
            "video: Generated video with audio"
          ]
        },
        "minComfyUIVersion": "0.30.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video"
          ],
          "model_options": {
            "GrokVideoNode": [
              "grok-imagine-video",
              "grok-imagine-video-1.5",
              "grok-imagine-video-1.5-lite"
            ]
          }
        }
      },
      {
        "name": "api_grok_reference_to_video",
        "title": "Grok: Reference to Video",
        "task": "Character Reference",
        "model": "Grok",
        "freshness": "established",
        "usage": 44,
        "recommend": "not_recommended",
        "description": "Generate a video from up to six reference images and a text prompt using the Grok API for fast, high-quality output. It ensures character or subject consistency across frames. Outputs a generated video; execution time depends on the third-party server response.",
        "io": {
          "inputs": [
            "image×2: Reference images for character or subject consistency",
            "text: Text prompt describing the video scene"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.18.2",
        "capabilities": {
          "workflow": [
            "api",
            "character-reference"
          ],
          "model_options": {
            "GrokVideoReferenceNode": [
              "grok-imagine-video-1.5",
              "grok-imagine-video"
            ]
          }
        }
      },
      {
        "name": "api_bytedance_image_to_video",
        "title": "Seedance 1.0 Pro: Image to Video",
        "task": "Image to Video",
        "model": "Seedance1.0 Pro",
        "freshness": "established",
        "usage": 603,
        "recommend": "high",
        "description": "This workflow transforms a static input image into a dynamic video using ByteDance’s Seedance 1.0 Pro via a third‑party API call. It accepts an image and generates a video with natural motion and consistent style, supporting model variants like seedance‑1‑0‑pro‑250528. Execution time depends on server‑side response.",
        "io": {
          "inputs": [
            "image: Input image to animate into a video"
          ],
          "outputs": [
            "video: Generated video with natural motion"
          ]
        },
        "minComfyUIVersion": "0.3.58",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video"
          ],
          "model_options": {
            "ByteDanceImageToVideoNode": [
              "seedance-1-5-pro-251215",
              "seedance-1-0-pro-250528",
              "seedance-1-0-pro-fast-251015"
            ]
          }
        }
      },
      {
        "name": "api_minimax_h3_max_r2v",
        "title": "MiniMax H3 Max: Reference to Video",
        "task": "Reference to Video",
        "model": "MiniMax H3",
        "freshness": "recent",
        "usage": 360,
        "recommend": "medium",
        "description": "Generate video from an image reference using MiniMax H3 Max, which animates your still frame while preserving its composition, visual identity, and aspect ratio. The model generates faster than real time with strong prompt adherence and polished cinematic motion. Ideal for bringing character art to life, animating product shots, and creating video from concept images.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.34.2",
        "capabilities": {
          "workflow": [
            "api",
            "reference-to-video"
          ]
        }
      },
      {
        "name": "video_minimax_h3_multiframe_reference",
        "title": "MiniMax H3: Multiframe Reference",
        "task": "Reference to Video",
        "model": "MiniMax H3",
        "freshness": "recent",
        "usage": 296,
        "recommend": "medium",
        "description": "Generate video from multiple still images with MiniMax H3, anchoring up to four reference frames anywhere along the timeline to control motion and composition. Ideal for creating video continuations from existing clips, forcing narrative beats at specific moments, and producing footage with synchronized native audio for cinematic or social media content.",
        "io": {
          "inputs": [
            "video: Input video"
          ],
          "outputs": [
            "video: Interpolated video"
          ]
        },
        "minComfyUIVersion": "0.34.0",
        "capabilities": {
          "workflow": [
            "reference-to-video",
            "int8"
          ]
        }
      },
      {
        "name": "api_minimax_h3_max_turbo_i2v",
        "title": "MiniMax H3 Max Turbo: Image to Video",
        "task": "Reference to Video",
        "model": "MiniMax H3",
        "freshness": "recent",
        "usage": 92,
        "recommend": "low",
        "description": "Generate video sequences from a starting image using MiniMax H3 Max Turbo, producing two video outputs from a single input frame. This rapid image-to-video workflow generates short clips in seconds, with optional prompt enhancement and 2K upscaling stages. Ideal for animating still photographs, creating quick motion previews for storyboards, and fast visual effects iteration.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.34.4",
        "capabilities": {
          "workflow": [
            "api",
            "reference-to-video"
          ]
        }
      },
      {
        "name": "api_minimax_h3_max_i2v",
        "title": "MiniMax H3 Max: Image to Video",
        "task": "Image to Video",
        "model": "MiniMax H3",
        "freshness": "recent",
        "usage": 231,
        "recommend": "medium",
        "description": "Generate video from a source image using MiniMax H3 Max, animating your still frame while preserving its composition, visual identity, and aspect ratio. The model delivers faster-than-real-time generation with strong prompt adherence and polished cinematic aesthetics. Ideal for bringing character art to life, animating product shots, and creating motion from concept images.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.34.3",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video"
          ]
        }
      },
      {
        "name": "video_minimax_h3_fun_controlnet_union",
        "title": "MiniMax H3 Fun ControlNet Union",
        "task": "Video Generation",
        "model": "MiniMax H3",
        "freshness": "recent",
        "usage": 154,
        "recommend": "low",
        "description": "Generate videos from a pose reference video using MiniMax-H3 with a ControlNet-Union checkpoint, preserving the input's motion and structure while applying your prompt. Ideal for character animation, motion transfer, and choreography visualization projects. This workflow is for pose control only",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.35.0",
        "capabilities": {
          "workflow": [
            "controlnet",
            "int8"
          ]
        }
      },
      {
        "name": "api_minimax_h3_r2v",
        "title": "MiniMax H3: Reference to Video",
        "task": "Reference to Video",
        "model": "MiniMax H3",
        "freshness": "current",
        "usage": 147,
        "recommend": "low",
        "description": "Generate video from a reference image or set of multimodal inputs using MiniMax H3, which interprets text, images, video, and audio together to produce coherent 5–15 second clips with native stereo audio. Input up to 2 images for first/last frame control, or up to 9 images, 3 video clips, and 3 audio clips in Omni Reference mode, and output a single video in your chosen aspect ratio. Ideal for commercial content creation, character-driven storytelling, and rapid visual concept validation.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.30.0",
        "capabilities": {
          "workflow": [
            "api",
            "reference-to-video",
            "image-to-video"
          ],
          "model_options": {
            "MinimaxHailuo03ReferenceNode": [
              "MiniMax H3"
            ]
          }
        }
      },
      {
        "name": "api_minimax_h3_max_t2v",
        "title": "MiniMax H3 Max: Text to Video",
        "task": "Text to Video",
        "model": "MiniMax H3",
        "freshness": "recent",
        "usage": 107,
        "recommend": "low",
        "description": "Generate polished, prompt-faithful video from text using MiniMax H3 Max, producing a single finished video clip with no input files required. Ideal for rapid concept exploration, cinematic text-to-video scenes, and high-throughput creative iteration where speed and visual quality matter.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.34.2",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-video"
          ]
        }
      },
      {
        "name": "api_minimax_h3_max_flf2v",
        "title": "MiniMax H3 Max: FLF2V",
        "task": "FLF2V",
        "model": "MiniMax H3",
        "freshness": "recent",
        "usage": 101,
        "recommend": "low",
        "description": "Generate polished, prompt-faithful videos faster than real time with MiniMax H3 Max, a post-trained model tuned for stronger prompt adherence and vivid cinematic aesthetics. This workflow supports text-to-video, image-to-video animation, and directed first-to-last-frame transitions, delivering clean compositions and cohesive lighting directly from the model. Ideal for rapid concept exploration, cinematic motion graphics, and high-throughput video comparison.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.34.2",
        "capabilities": {
          "workflow": [
            "api",
            "flf2v"
          ]
        }
      },
      {
        "name": "api_kling_v3_video",
        "title": "Kling 3.0: Video Generation",
        "task": "Image to Video",
        "model": "Kling 3.0",
        "freshness": "established",
        "usage": 517,
        "recommend": "high",
        "description": "Generate up to 15-second video sequences from text prompts or optional reference images using the Kling 3.0 frontier model via API. Supports native audio, multi-shot compositions, and complex camera movements. Input an image and/or text prompt; outputs a generated video. This workflow calls a third-party API; execution time depends on server-side response.",
        "io": {
          "inputs": [
            "image: Optional reference image to start from",
            "text: Text prompt describing the video content"
          ],
          "outputs": [
            "video: Generated video sequence"
          ]
        },
        "minComfyUIVersion": "0.13.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video"
          ],
          "model_options": {
            "KlingVideoNode": [
              "kling-v3",
              "kling-3.0-turbo"
            ]
          }
        }
      },
      {
        "name": "api_kling_v3_flf2v",
        "title": "Kling 3.0: First Last Frame to Video",
        "task": "FLF2V",
        "model": "Kling 3.0",
        "freshness": "established",
        "usage": 249,
        "recommend": "medium",
        "description": "This Kling 3.0 First-Last Frame to Video workflow takes a beginning and ending image, then generates a continuous video sequence that smoothly interpolates between them. It supports multi-shot generation, precise element control, and multilingual prompts at up to 4K resolution. As a third-party API workflow, execution time depends on server-side processing.",
        "io": {
          "inputs": [
            "image×2: First frame and last frame images defining the video start and end"
          ],
          "outputs": [
            "video: Generated video sequence between the two frames"
          ]
        },
        "minComfyUIVersion": "0.13.0",
        "capabilities": {
          "workflow": [
            "api",
            "flf2v"
          ],
          "model_options": {
            "KlingFirstLastFrameNode": [
              "kling-v3"
            ]
          }
        }
      },
      {
        "name": "api_kling_motion_control3",
        "title": "Kling 3.0: Motion Control",
        "task": "Motion Control",
        "model": "Kling 3.0",
        "freshness": "established",
        "usage": 152,
        "recommend": "low",
        "description": "Transfer motion from a reference video to a character image using Kling 3.0's motion control API. The latest frontier model from Kling supports durations up to 15 seconds with native audio. This workflow requires a character image and a motion reference video, and execution depends on server-side response.",
        "io": {
          "inputs": [
            "image: Character image to animate",
            "video: Motion reference video to transfer actions from"
          ],
          "outputs": [
            "video: Generated video with transferred motion"
          ]
        },
        "minComfyUIVersion": "0.7.0",
        "capabilities": {
          "workflow": [
            "api",
            "motion-control"
          ],
          "model_options": {
            "KlingMotionControl": [
              "kling-v3",
              "kling-v2-6"
            ]
          }
        }
      },
      {
        "name": "api_pruna_p_video_i2v",
        "title": "Pruna P-Video 2: Image to Video",
        "task": "Image to Video",
        "model": "P-Video-2",
        "freshness": "recent",
        "usage": 84,
        "recommend": "low",
        "description": "Turn a single image into video with Pruna P-Video-2. As Pruna's quality-focused video model, its image-to-video mode excels at identity consistency and subtle, natural motion — 720p/1080p, 1-20s duration, plus Draft mode for fast previews. Ideal for image animation, product motion, and poster cinemagraphs.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.36.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video"
          ]
        }
      },
      {
        "name": "video_wan2_2_14B_t2v",
        "title": "Wan 2.2 14B Text to Video",
        "task": "Text to Video",
        "model": "Wan2.2",
        "freshness": "established",
        "usage": 291,
        "recommend": "medium",
        "description": "Generate high-quality videos from text prompts with cinematic aesthetic control and dynamic motion generation using Wan 2.2.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.16.4",
        "capabilities": {
          "workflow": [
            "text-to-video"
          ]
        }
      },
      {
        "name": "video_wan_vace_inpainting",
        "title": "Wan2.1 VACE Inpainting",
        "task": "Image Inpainting",
        "model": "Wan2.1",
        "freshness": "established",
        "usage": 206,
        "recommend": "medium",
        "description": "Edit specific regions in videos while preserving surrounding content. Great for object removal or replacement.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.20.0",
        "capabilities": {
          "workflow": [
            "inpainting"
          ]
        }
      },
      {
        "name": "video_wan2_2_14B_flf2v",
        "title": "Wan 2.2 14B First-Last Frame to Video",
        "task": "FLF2V",
        "model": "Wan2.2",
        "freshness": "established",
        "usage": 347,
        "recommend": "medium",
        "description": "Generate smooth video transitions by defining start and end frames.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.31",
        "capabilities": {
          "workflow": [
            "flf2v"
          ]
        }
      },
      {
        "name": "video_wan2_2_5B_ti2v",
        "title": "Wan 2.2 5B Video Generation",
        "task": "Text to Video",
        "model": "Wan2.2",
        "freshness": "established",
        "usage": 334,
        "recommend": "medium",
        "description": "Fast text-to-video and image-to-video generation with 5B parameters. Optimized for rapid prototyping and creative exploration.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.3.46",
        "capabilities": {
          "workflow": [
            "text-to-video"
          ]
        }
      },
      {
        "name": "video_wan_vace_14B_v2v",
        "title": "Wan2.1 VACE Control Video",
        "task": "Video to Video",
        "model": "Wan2.1",
        "freshness": "established",
        "usage": 156,
        "recommend": "low",
        "description": "Generate videos by controlling input videos and reference images using Wan VACE.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.31",
        "capabilities": {
          "workflow": [
            "video-to-video"
          ]
        }
      },
      {
        "name": "image_to_video_wan",
        "title": "Wan 2.1 Image to Video",
        "task": "Text to Video",
        "model": "Wan2.1",
        "freshness": "established",
        "usage": 100,
        "recommend": "low",
        "description": "Generate videos from images using Wan 2.1.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.3.31",
        "capabilities": {
          "workflow": [
            "text-to-video"
          ]
        }
      },
      {
        "name": "video_wan2_2_14B_fun_control",
        "title": "Wan 2.2 14B Fun Control",
        "task": "Video to Video",
        "model": "Wan2.2",
        "freshness": "established",
        "usage": 41,
        "recommend": "not_recommended",
        "description": "Generate videos guided by pose, depth, and edge controls using Wan 2.2 Fun Control.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.50",
        "capabilities": {
          "workflow": [
            "video-to-video",
            "controlnet"
          ]
        }
      },
      {
        "name": "video_bernini_r_video_editing",
        "title": "Bernini-R: Video Edit",
        "task": "Video Edit",
        "model": "Bernini-R",
        "freshness": "current",
        "usage": 287,
        "recommend": "medium",
        "description": "Generate an edited video with consistent relighting using Bernini-R. Ideal for portrait and product relighting, e-commerce catalog photography, and creating consistent lighting across photo sets.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.25.0",
        "capabilities": {
          "workflow": [
            "video-edit"
          ]
        }
      },
      {
        "name": "api_google_gemini_omni_flash_1_1_edit",
        "title": "Gemini Omni 1.1 Flash: Video Edit",
        "task": "Video Edit",
        "model": "Gemini Omni 1.1 Flash",
        "freshness": "recent",
        "usage": 281,
        "recommend": "medium",
        "description": "Edit videos through natural language conversation with Gemini Omni 1.1 Flash. Upload a clip (10 seconds or less) and describe the change: swap backgrounds, restyle scenes, add or remove elements, and the model applies the edit while preserving everything you did not mention. Add \\\"Keep everything else the same\\\" for maximum consistency; simple prompts work best. Outputs video with native audio at 16:9 or 9:16, with 1080p/4K output available. Ideal for remixes, style changes, and iterative refinement without re-generating from scratch.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.34.2",
        "capabilities": {
          "workflow": [
            "api",
            "video-edit"
          ],
          "model_options": {
            "GeminiVideoOmniV2": [
              "Omni Flash 1.1",
              "Omni Flash"
            ]
          }
        }
      },
      {
        "name": "api_google_gemini_omni_flash_1_1_i2v",
        "title": "Gemini Omni 1.1 Flash: Image to Video",
        "task": "Image to Video",
        "model": "Gemini Omni 1.1 Flash",
        "freshness": "recent",
        "usage": 128,
        "recommend": "low",
        "description": "Generate videos from text prompts or reference images using Gemini Omni 1.1 Flash, a multimodal model that handles text-to-video, image-to-video, keyframe interpolation, and iterative video editing with natural language. Outputs a single MP4 video with optional audio, supporting aspect ratios, resolutions up to 4K, and scene extension. Ideal for cinematic storytelling, product visualization, and rapid video prototyping.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.34.2",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video"
          ],
          "model_options": {
            "GeminiVideoOmniV2": [
              "Omni Flash 1.1",
              "Omni Flash"
            ]
          }
        }
      },
      {
        "name": "api_google_gemini_omni_flash_1_1_r2v",
        "title": "Gemini Omni 1.1 Flash: Reference to Video",
        "task": "Reference to Video",
        "model": "Gemini Omni 1.1 Flash",
        "freshness": "recent",
        "usage": 143,
        "recommend": "low",
        "description": "Generate video that incorporates specific subjects from multiple reference images using Gemini Omni 1.1 Flash's reference mode: characters, products, and objects from your images appear in the scene while the image itself is never shown as a frame. Bind each image to a role with tags like <IMAGE_REF_0>, combine character references with style references, and direct the scene in plain language with native audio. Ideal for character-driven stories, product placement, and brand-consistent content.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.34.2",
        "capabilities": {
          "workflow": [
            "api",
            "reference-to-video"
          ],
          "model_options": {
            "GeminiVideoOmniV2": [
              "Omni Flash 1.1",
              "Omni Flash"
            ]
          }
        }
      },
      {
        "name": "api_google_gemini_omni_flash_1_1_t2v",
        "title": "Gemini Omni 1.1 Flash: Text to Video",
        "task": "Text to Video",
        "model": "Gemini Omni 1.1 Flash",
        "freshness": "recent",
        "usage": 78,
        "recommend": "low",
        "description": "Generate cinematic videos with audio from text prompts or reference images using Gemini Omni 1.1 Flash, supporting aspect ratio, resolution, and keyframe interpolation controls. Ideal for text-to-video storytelling, animating product shots or illustrations, and iterative conversational video editing.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.34.2",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-video"
          ],
          "model_options": {
            "GeminiVideoOmniV2": [
              "Omni Flash 1.1",
              "Omni Flash"
            ]
          }
        }
      },
      {
        "name": "video_ltx2_3_flf2v",
        "title": "LTX-2.3: FLF2V",
        "task": "FLF2V",
        "model": "LTX-2.3",
        "freshness": "established",
        "usage": 150,
        "recommend": "low",
        "description": "Upload a starting image to generate a video sequence. The workflow creates a short video with consistent style and motion, beginning with your provided frame.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.16.4",
        "capabilities": {
          "workflow": [
            "flf2v"
          ]
        }
      },
      {
        "name": "video_ltx2_3_id_lora",
        "title": "LTX-2.3: ID LoRA",
        "task": "Lip Sync",
        "model": "LTX-2.3",
        "freshness": "established",
        "usage": 117,
        "recommend": "low",
        "description": "Generate personalized videos with synchronized audio from a text prompt, reference image, and short audio clip. Use ID-LoRA to adapt a person's appearance and voice in a single generative model.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.19.0",
        "capabilities": {
          "workflow": [
            "lip-sync",
            "voice-cloning"
          ]
        }
      },
      {
        "name": "video_ltx2_3_ic_lora",
        "title": "LTX 2.3 IC-LoRA Union Control",
        "task": "Video Generation",
        "model": "LTX-2.3",
        "freshness": "established",
        "usage": 85,
        "recommend": "low",
        "description": "Generate LTX 2.3 videos with IC-LoRA using aligned control inputs like depth, pose, or edges.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.22.0",
        "capabilities": {
          "workflow": [
            "controlnet"
          ]
        }
      },
      {
        "name": "api_kling_o3_video_edit",
        "title": "Kling O3: Video Edit",
        "task": "Video Edit",
        "model": "Kling O3",
        "freshness": "established",
        "usage": 204,
        "recommend": "medium",
        "description": "This workflow edits a source video using Kling O3’s video generation and image editing features. It takes a video and a reference image as inputs to produce a professionally edited output video with precise subject and scene adjustments. Execution is handled by a third-party API.",
        "io": {
          "inputs": [
            "video: Source video to edit",
            "image: Reference image for subject or style guidance"
          ],
          "outputs": [
            "video: Edited output video"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "video-edit",
            "api"
          ]
        }
      },
      {
        "name": "api_kling4_i2v",
        "title": "Kling 4.0: Image to Video",
        "task": "Image to Video",
        "model": "Kling 4.0",
        "freshness": "new",
        "usage": 173,
        "recommend": "medium",
        "description": "Animate a single still into a five-second clip with Kling 4.0, cutting three times inside one generation. The source image locks the subject and styling while the default prompt drives the camera moves and a beat-timed edit, and the model also produces stereo audio. Ideal for fashion, editorial, and product motion where the input frame defines the look.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.42.6",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video"
          ]
        }
      },
      {
        "name": "api_kling4_t2v",
        "title": "Kling 4.0: Text to Video",
        "task": "Text to Video",
        "model": "Kling 4.0",
        "freshness": "new",
        "usage": 22,
        "recommend": "medium",
        "description": "Generate a ten-second cinematic clip from text alone with Kling 4.0, holding one single continuous shot through a full high-speed chase while also producing native stereo audio. Ideal for short-form spectacle, concept pitches, and motion references where a single unbroken take carries the idea.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.42.6",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-video"
          ]
        }
      },
      {
        "name": "api_kling4_video_edit",
        "title": "Kling 4.0: Video Edit",
        "task": "Video Edit",
        "model": "Kling 4.0",
        "freshness": "new",
        "usage": 108,
        "recommend": "medium",
        "description": "Edit an existing clip in one pass with Kling 4.0: name the change and list what must survive, and the model rewrites only that part while keeping the subject, costume, shot count, camera work and timing intact. Output inherits the input's length and aspect ratio, and stereo audio is regenerated to match the new scene. Ideal for swapping environments, wardrobe or style on footage that is already cut.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.42.6",
        "capabilities": {
          "workflow": [
            "video-edit",
            "api"
          ]
        }
      },
      {
        "name": "api_kling4_r2v",
        "title": "Kling 4.0: Reference to Video",
        "task": "Reference to Video",
        "model": "Kling 4.0",
        "freshness": "new",
        "usage": 147,
        "recommend": "medium",
        "description": "Hold a character and a location from two reference images and shoot a three-shot fashion fragment between them: an extreme close-up with a snap zoom, a mid-hall turn with a whip-pan, and a 120-degree orbit under the arched window. Each cut lands on a new position inside the same room, with two hard cuts in five seconds and stereo audio generated for the space.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.42.6",
        "capabilities": {
          "workflow": [
            "api",
            "reference-to-video"
          ]
        }
      },
      {
        "name": "api_kling_o3_i2v",
        "title": "Kling O3: Reference to Video",
        "task": "Character Reference",
        "model": "Kling O3",
        "freshness": "established",
        "usage": 110,
        "recommend": "low",
        "description": "Upload a reference image and optional text prompt to generate a character-consistent video using Kling O3. This image-to-video workflow supports native audio output and precise storyboard control via a third-party API, so execution time depends on server-side response.",
        "io": {
          "inputs": [
            "image: Reference image of the subject or scene to animate"
          ],
          "outputs": [
            "video: Generated character-consistent video with optional audio"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "api",
            "character-reference",
            "image-to-video"
          ]
        }
      },
      {
        "name": "api_grok_imagine_video_1_5_lite_i2v",
        "title": "Grok Imagine Video 1.5 Lite: Image to Video",
        "task": "Image to Video",
        "model": "Grok Imagine Video",
        "freshness": "recent",
        "usage": 263,
        "recommend": "medium",
        "description": "Generate short videos from a single still image using Grok Imagine Video 1.5 Lite, a faster, lower-cost model that turns one image into a video clip up to 15 seconds at resolutions up to 1080p. Ideal for rapid social content, storyboard previews, and high-volume creative iteration where speed and cost matter more than maximum fidelity.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.38.2",
        "capabilities": {
          "workflow": [
            "image-to-video",
            "api"
          ],
          "model_options": {
            "GrokVideoNode": [
              "grok-imagine-video",
              "grok-imagine-video-1.5",
              "grok-imagine-video-1.5-lite"
            ]
          }
        }
      },
      {
        "name": "api_grok_imagine_video_1_5_lite_t2v",
        "title": "Grok Imagine Video 1.5 Lite: Text to Video",
        "task": "Text to Video",
        "model": "Grok Imagine Video",
        "freshness": "recent",
        "usage": 31,
        "recommend": "not_recommended",
        "description": "Generate a video of up to 15 seconds from a text prompt with Grok Imagine Video 1.5 Lite, the distilled and faster version of Imagine Video 1.5, with native audio and 480p, 720p or 1080p output in any of seven aspect ratios. No input files are required and one video is output. Ideal for social clips, ad mockups, and fast iteration where a full-quality render is not worth the wait.\n",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.38.2",
        "capabilities": {
          "workflow": [
            "text-to-video",
            "api"
          ],
          "model_options": {
            "GrokVideoNode": [
              "grok-imagine-video",
              "grok-imagine-video-1.5",
              "grok-imagine-video-1.5-lite"
            ]
          }
        }
      },
      {
        "name": "api_seedance2_5_flf2v",
        "title": "Seedance 2.5: FLF2V",
        "task": "FLF2V",
        "model": "Seedance 2.5",
        "freshness": "recent",
        "usage": 52,
        "recommend": "low",
        "description": "Generate a smooth video transition between a user-provided start frame and end frame with Seedance 2.5, which automatically interpolates the motion, camera movement, and visual evolution while locking both endpoints. Ideal for shot transitions, storyboard previz, and before-and-after scene showcases like empty-to-packed or day-to-night shifts.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.31.0",
        "capabilities": {
          "workflow": [
            "api",
            "flf2v"
          ],
          "model_options": {
            "ByteDance2FirstLastFrameNode": [
              "Seedance 2.5",
              "Seedance 2.5 Draft",
              "Seedance 2.0",
              "Seedance 2.0 Fast",
              "Seedance 2.0 Mini"
            ]
          }
        }
      },
      {
        "name": "api_seedance2_5_draft_t2v",
        "title": "Seedance2.5 Draft: Text to Video",
        "task": "Text to Video",
        "model": "Seedance 2.5",
        "freshness": "recent",
        "usage": 51,
        "recommend": "low",
        "description": "Generate draft preview videos from text prompts using Seedance 2.5's two-phase draft mode: stage one turns your prompt into a fast, low-cost 480p preview plus a reusable draft task ID, and stage two renders the 1080p final from that ID. Ideal for storyboard testing, rapid creative iteration, and budget-conscious production workflows.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.37.3",
        "capabilities": {
          "workflow": [
            "api"
          ],
          "model_options": {
            "ByteDance2TextToVideoNode": [
              "Seedance 2.5",
              "Seedance 2.5 Draft",
              "Seedance 2.0",
              "Seedance 2.0 Fast",
              "Seedance 2.0 Mini"
            ]
          }
        }
      },
      {
        "name": "video_ltx2_i2v",
        "title": "LTX-2: Image to Video",
        "task": "Image to Video",
        "model": "LTX-2",
        "freshness": "established",
        "usage": 106,
        "recommend": "low",
        "description": "Transform static images into dynamic videos with synchronized audio-video generation using LTX-2. Features expressive lip sync, natural motion, and efficient performance.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "image-to-video"
          ]
        }
      },
      {
        "name": "video_ltx2_i2v_distilled",
        "title": "LTX-2 Image to Video (Distilled)",
        "task": "Image to Video",
        "model": "LTX-2",
        "freshness": "established",
        "usage": 23,
        "recommend": "not_recommended",
        "description": "Transform static images into dynamic videos with synchronized audio-video generation using LTX-2 distilled model. Optimized for faster generation while maintaining quality. Features expressive lip sync, natural motion, and improved speed.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "image-to-video"
          ]
        }
      },
      {
        "name": "video_ltx2_t2v",
        "title": "LTX-2 Text to Video",
        "task": "Text to Video",
        "model": "LTX-2",
        "freshness": "established",
        "usage": 20,
        "recommend": "not_recommended",
        "description": "Generate high-quality videos from text prompts with synchronized audio-video generation using LTX-2. Features expressive lip sync, dynamic motion generation, and efficient performance.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "text-to-video"
          ]
        }
      },
      {
        "name": "video_ltx2_canny_to_video",
        "title": "LTX-2 Canny to Video",
        "task": "Video Generation",
        "model": "LTX-2",
        "freshness": "established",
        "usage": 14,
        "recommend": "not_recommended",
        "description": "Generate high-quality videos from edge detection (Canny) guidance with synchronized audio-video generation using LTX-2. Features precise edge control, structural consistency, and efficient performance.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "controlnet"
          ]
        }
      },
      {
        "name": "api_flux_video_edit",
        "title": "Flux Video Edit",
        "task": "Video Edit",
        "model": "FLUX",
        "freshness": "recent",
        "usage": 54,
        "recommend": "low",
        "description": "Edit and transform images with FLUX using simple text prompts. Change characters, objects, backgrounds, styles, or entire scenes while preserving the key details and overall consistency of the original image.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.36.0",
        "capabilities": {
          "workflow": [
            "api",
            "video-edit"
          ]
        }
      },
      {
        "name": "api_wan3_0_t2v",
        "title": "Wan3.0: Text to Video",
        "task": "Text to Video",
        "model": "Wan3.0",
        "freshness": "recent",
        "usage": 90,
        "recommend": "low",
        "description": "Generate videos up to 30 seconds directly from text prompts using Wan3.0-Video, producing a single video output with lifelike visuals and synchronized sound. This workflow transforms written descriptions into immersive, production-ready footage with strong character consistency and cinematic quality. Ideal for cinematic storytelling, character-driven animations, and quick multimedia content creation.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.33.4",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-video"
          ],
          "model_options": {
            "Wan3ReferenceToVideoApi": [
              "wan3.0-video",
              "wan3.0-video-prime"
            ]
          }
        }
      },
      {
        "name": "api_heygen_video_1_r2v",
        "title": "HeyGen Video 1.0: Reference to Video",
        "task": "Reference to Video",
        "model": "HeyGen",
        "freshness": "recent",
        "usage": 121,
        "recommend": "low",
        "description": "Generate a 5-15 second video with dialogue and sound from a prompt plus up to nine images, three videos and three audio clips using HeyGen Video 1.0, addressed in the prompt as @Image1, @Video1 and @Audio1. One video is output, with the people, products and places you supply driving the shot. Ideal for product demos, UGC-style clips that keep one presenter or product consistent, and placing your product into a scene of your own.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.38.1",
        "capabilities": {
          "workflow": [
            "api",
            "reference-to-video"
          ],
          "model_options": {
            "HeyGenReferenceToVideoNode": [
              "heygen-video-1"
            ]
          }
        }
      },
      {
        "name": "api_heygen_video_1_t2v",
        "title": "HeyGen Video 1.0: Text to Video",
        "task": "Text to Video",
        "model": "HeyGen",
        "freshness": "recent",
        "usage": 43,
        "recommend": "not_recommended",
        "description": "Generate a 5-15 second clip with synchronized dialogue and sound from a text prompt using HeyGen Video 1.0, at 480p or 768p in any of six aspect ratios, with no input files required. One video is output. Ideal for UGC-style talking clips, product explainers, and short social ads where the spoken line and the ambience have to match the scene.\n",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.38.1",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-video"
          ],
          "model_options": {
            "HeyGenReferenceToVideoNode": [
              "heygen-video-1"
            ]
          }
        }
      },
      {
        "name": "api_heygen_talking_photo",
        "title": "HeyGen: Talking Photo",
        "task": "Image to Video",
        "model": "HeyGen",
        "freshness": "current",
        "usage": 36,
        "recommend": "not_recommended",
        "description": "Generate a talking-head video from a single photograph using HeyGen's API, with realistic lip-synced speech, facial expressions, and head movements. This image-to-video workflow takes a text prompt as input and produces an animated avatar output, ideal for digital presentations and personalized video messages.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.28.1",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video",
            "lip-sync"
          ]
        }
      },
      {
        "name": "api_veed_fabric_1_0",
        "title": "VEED Fabric 1.0: Talking Photo",
        "task": "Image to Video",
        "model": "VEED",
        "freshness": "new",
        "usage": 40,
        "recommend": "medium",
        "description": "Turn a still photo and a speech track into a talking, lip-synced video with VEED Fabric 1.0. The face in the image speaks the audio, the clip keeps the image's aspect ratio and runs as long as the audio, up to 300 seconds at 480p or 720p. Ideal for avatars, mascots and localized spokesperson clips driven by your own voice recording or a TTS node.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.43.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video",
            "lip-sync"
          ]
        }
      },
      {
        "name": "api_veed_lipsync_2_0",
        "title": "VEED Lipsync 2.0: Lip Sync",
        "task": "Lip Sync",
        "model": "VEED",
        "freshness": "new",
        "usage": 164,
        "recommend": "medium",
        "description": "Re-sync a speaker's mouth in footage you already shot to a new speech track with VEED Lipsync 2.0. The person on camera speaks the new audio while the rest of the shot stays as recorded; the output keeps the video's resolution at 25 fps and runs as long as the audio, up to 600 seconds. If the clip is shorter than the speech it plays forward and backward until the audio ends. Ideal for dubbing a line, swapping a script without re-shooting, or making localized versions of existing footage.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.43.0",
        "capabilities": {
          "workflow": [
            "api",
            "lip-sync",
            "video-edit"
          ]
        }
      },
      {
        "name": "api_seedance2_0_mini_r2v",
        "title": "Seedance 2.0 Mini: Reference to Video",
        "task": "Image to Video",
        "model": "Seedance 2.0",
        "freshness": "current",
        "usage": 100,
        "recommend": "low",
        "description": "Generate cinematic video from two reference images using Seedance 2.0 Mini, producing a single video output with AI-driven camera motion and consistent character generation. Ideal for storytelling, branded content, and short film creation where shot composition and visual continuity are essential.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.26.2",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video",
            "reference-to-video"
          ],
          "model_options": {
            "ByteDance2ReferenceNodeV2": [
              "Seedance 2.5",
              "Seedance 2.5 Draft",
              "Seedance 2.0",
              "Seedance 2.0 Fast",
              "Seedance 2.0 Mini"
            ]
          }
        }
      },
      {
        "name": "api_seedance2_0_r2v_4k",
        "title": "Seedance2.0 4K: Reference to Video",
        "task": "Image to Video",
        "model": "Seedance 2.0",
        "freshness": "current",
        "usage": 57,
        "recommend": "low",
        "description": "Generate a reference-based video from a single input image using Seedance 2.0, producing native 4K output with 10-bit color depth and clean fine detail retention. The workflow takes 1 image and outputs 1 video. Ideal for high-resolution motion graphics, cinematic VFX shots requiring downstream grading, and professional video production where 4K finishing is essential.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.26.2",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video",
            "reference-to-video"
          ],
          "model_options": {
            "ByteDance2ReferenceNodeV2": [
              "Seedance 2.5",
              "Seedance 2.5 Draft",
              "Seedance 2.0",
              "Seedance 2.0 Fast",
              "Seedance 2.0 Mini"
            ]
          }
        }
      },
      {
        "name": "template_seedance2_storyboard_to_video",
        "title": "Storyboard To Video - Seedance 2.0",
        "task": "Image to Video",
        "model": "Seedance 2.0",
        "freshness": "current",
        "usage": 67,
        "recommend": "low",
        "description": "Generate an 8-panel storyboard from a simple text prompt, then use that storyboard as the foundation for Seedance 2.0. The storyboard provides a visual blueprint for the sequence, allowing you to guide the video with more control over the action, camera angles, composition, and overall storytelling.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.21.1",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video"
          ]
        }
      },
      {
        "name": "api_ltx2_5_i2v",
        "title": "LTX-2.5 (Pro): Image to Video",
        "task": "Image to Video",
        "model": "LTX-2.5",
        "freshness": "recent",
        "usage": 43,
        "recommend": "not_recommended",
        "description": "Generate video from a single image using LTX-2.5's fast distilled model, delivering cinematic multi-shot sequences with consistent characters, lighting, and style across cuts. The workflow outputs a video file with native multishot support and automatic duration prediction. Ideal for rapid storyboarding, social media content creation, and pre-visualization in commercial production pipelines.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.32.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video"
          ]
        }
      },
      {
        "name": "api_ltx2_5_flf2v",
        "title": "LTX-2.5 (Pro): FLF2V",
        "task": "FLF2V",
        "model": "LTX-2.5",
        "freshness": "recent",
        "usage": 26,
        "recommend": "not_recommended",
        "description": "Generate a video from two reference images using LTX-2.5, blending the content of both frames into a single animated sequence. Ideal for creating transition shots, morphing between two subjects or scenes, and building multi-shot narratives from still references.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.32.0",
        "capabilities": {
          "workflow": [
            "api",
            "flf2v"
          ]
        }
      },
      {
        "name": "video_hunyuan_video_1.5_720p_i2v",
        "title": "Hunyuan Video 1.5 Image to Video",
        "task": "Image to Video",
        "model": "Hunyuan Video",
        "freshness": "established",
        "usage": 114,
        "recommend": "low",
        "description": "Animate still images into dynamic videos with precise motion and camera control. Maintains visual consistency while bringing photos and illustrations to life with smooth, natural movements.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.3.71",
        "capabilities": {
          "workflow": [
            "image-to-video"
          ]
        }
      },
      {
        "name": "video_hunyuan_video_1.5_720p_t2v",
        "title": "Hunyuan Video 1.5 Text to Video",
        "task": "Text to Video",
        "model": "Hunyuan Video",
        "freshness": "established",
        "usage": 19,
        "recommend": "not_recommended",
        "description": "Generate high-quality 720p videos from text prompts with cinematic camera control, emotional expressions, and physics simulation. Supports multiple styles including realistic, anime, and 3D with text rendering.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.3.71",
        "capabilities": {
          "workflow": [
            "text-to-video"
          ]
        }
      },
      {
        "name": "api_bfl_flux3_i2v",
        "title": "FLUX 3 Video: Image to Video",
        "task": "Image to Video",
        "model": "FLUX 3 Video",
        "freshness": "current",
        "usage": 23,
        "recommend": "not_recommended",
        "description": "Generate full videos from a single text prompt using FLUX 3, which automatically expands your description into a cohesive scene with synchronized audio, motion, and physics. Outputs one 720p, 24fps MP4 video (5-20 seconds) with support for multi-shot structures and aspect ratios from 9:16 to 21:9. Ideal for product photography, filmic ad campaigns, and dynamic motion graphics.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.30.2",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video"
          ]
        }
      },
      {
        "name": "api_wan2_7_video_edit",
        "title": "Wan2.7: Video Edit",
        "task": "Video Edit",
        "model": "Wan2.7",
        "freshness": "established",
        "usage": 67,
        "recommend": "low",
        "description": "Edit and replace characters or scenes in a video using the Wan2.7 model. Upload a source video and two reference images, then provide text instructions describing the edits. The workflow outputs the edited video and relies on a third‑party API execution; response time depends on the server.",
        "io": {
          "inputs": [
            "video: Source video to edit",
            "image×2: Reference images for character or scene replacement",
            "text: Text instructions describing the edits"
          ],
          "outputs": [
            "video: Edited output video"
          ]
        },
        "minComfyUIVersion": "0.18.4",
        "capabilities": {
          "workflow": [
            "api",
            "video-edit"
          ],
          "model_options": {
            "Wan2VideoEditApi": [
              "wan2.7-videoedit"
            ]
          }
        }
      },
      {
        "name": "api_wan2_7_i2v",
        "title": "Wan2.7: Image to Video",
        "task": "Image to Video",
        "model": "Wan2.7",
        "freshness": "established",
        "usage": 64,
        "recommend": "low",
        "description": "This workflow uses the Wan2.7 Image-to-Video API, a medium-quality model from Alibaba, to animate a static image based on a text prompt. It outputs a video with natural motion at up to 1080p resolution. Since it calls a third-party API, execution time depends on server-side response.",
        "io": {
          "inputs": [
            "image: Input image to animate",
            "text: Text prompt describing the desired motion"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.18.4",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video"
          ],
          "model_options": {
            "Wan2ImageToVideoApi": [
              "wan2.7-i2v"
            ]
          }
        }
      },
      {
        "name": "api_sync_so_lip_sync_video",
        "title": "Sync 3: Video Lip-Sync",
        "task": "Lip Sync",
        "model": "Sync 3",
        "freshness": "current",
        "usage": 64,
        "recommend": "low",
        "description": "Sync 3 creates lip-synced videos by taking a video and an audio input, then adjusting the speaker's mouth movements to match the new audio track. This api-ready workflow outputs a single synced video, ideal for dubbing, dialogue correction, or animating a static portrait to speak.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.28.0",
        "capabilities": {
          "workflow": [
            "api",
            "lip-sync"
          ],
          "model_options": {
            "SyncLipSyncNode": [
              "sync-3"
            ]
          }
        }
      },
      {
        "name": "templates-6-key-frames",
        "title": "Multi-Keyframe Video Stitching",
        "task": "FLF2V",
        "model": "Wan2.2",
        "freshness": "established",
        "usage": 44,
        "recommend": "not_recommended",
        "description": "Creates a smooth video using 6 key frames.It auto fills in the motion between frames and stitches the segments together seamlessly.",
        "io": {
          "inputs": [
            "video: Input video"
          ],
          "outputs": [
            "video: Interpolated video"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "flf2v",
            "image-to-video"
          ]
        }
      },
      {
        "name": "video_wan2_2_14B_s2v",
        "title": "Wan2.2-S2V Audio-Driven Video Generation",
        "task": "Video Generation",
        "model": "Wan2.2",
        "freshness": "established",
        "usage": 98,
        "recommend": "low",
        "description": "Transform static images and audio into dynamic videos with perfect synchronization and minute-level generation.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.55",
        "capabilities": {
          "workflow": [
            "audio-to-video"
          ]
        }
      },
      {
        "name": "text_to_video_wan",
        "title": "Wan 2.1 Text to Video",
        "task": "Text to Video",
        "model": "Wan2.1",
        "freshness": "established",
        "usage": 52,
        "recommend": "low",
        "description": "Wan 2.1 Text to Video generates a video from a text prompt using the fast, low-quality Wan2.1 legacy model with video LoRA support. The workflow runs on Comfy Cloud, processing the input text to produce a video output.",
        "io": {
          "inputs": [
            "text: Text prompt describing the video"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.3.31",
        "capabilities": {
          "workflow": [
            "text-to-video"
          ]
        }
      },
      {
        "name": "video_wan_vace_outpainting",
        "title": "Wan2.1 VACE Outpainting",
        "task": "Video Generation",
        "model": "Wan2.1",
        "freshness": "established",
        "usage": 40,
        "recommend": "not_recommended",
        "description": "Generate extended videos by expanding video size using Wan VACE outpainting.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.31",
        "capabilities": {
          "workflow": [
            "outpainting"
          ]
        }
      },
      {
        "name": "video_causal_forcing_i2v",
        "title": "Causal Forcing: Image to Video",
        "task": "Image to Video",
        "model": "Wan2.1",
        "freshness": "established",
        "usage": 52,
        "recommend": "low",
        "description": "Input a text prompt and select a model checkpoint. Generate a high-quality video using Causal Forcing or Causal Forcing++ with 1 to 4 inference steps.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.21.0",
        "capabilities": {
          "workflow": [
            "image-to-video"
          ]
        }
      },
      {
        "name": "video_wan_vace_14B_ref2v",
        "title": "Wan2.1 VACE Reference to Video",
        "task": "Image to Video",
        "model": "Wan2.1",
        "freshness": "established",
        "usage": 36,
        "recommend": "not_recommended",
        "description": "Create videos that match the style and content of a reference image. Perfect for style-consistent video generation.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "image-to-video"
          ]
        }
      },
      {
        "name": "wan2.1_fun_control",
        "title": "Wan 2.1 ControlNet",
        "task": "Video to Video",
        "model": "Wan2.1",
        "freshness": "established",
        "usage": 17,
        "recommend": "not_recommended",
        "description": "Generate videos guided by pose, depth, and edge controls using Wan 2.1 ControlNet.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.31",
        "capabilities": {
          "workflow": [
            "video-to-video"
          ]
        }
      },
      {
        "name": "video_wan_vace_flf2v",
        "title": "Wan2.1 VACE First-Last Frame",
        "task": "FLF2V",
        "model": "Wan2.1",
        "freshness": "established",
        "usage": 15,
        "recommend": "not_recommended",
        "description": "Generate smooth video transitions by defining start and end frames. Supports custom keyframe sequences.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.16.4",
        "capabilities": {
          "workflow": [
            "flf2v"
          ]
        }
      },
      {
        "name": "video_wan_dancer",
        "title": "Wan Dancer: Music to Dance",
        "task": "Video Generation",
        "model": "Wan",
        "freshness": "current",
        "usage": 32,
        "recommend": "not_recommended",
        "description": "Generate minute-scale coherent dance videos from music using a hierarchical Wan Dancer framework. Input a music file and reference image to produce high-resolution dance videos with global structure and temporal continuity.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.21.0",
        "capabilities": {
          "workflow": [
            "audio-to-video"
          ]
        }
      },
      {
        "name": "api_wan2_6_i2v",
        "title": "Wan2.6: Image to Video",
        "task": "Image to Video",
        "model": "Wan2.6",
        "freshness": "established",
        "usage": 16,
        "recommend": "not_recommended",
        "description": "This workflow turns a reference image into a high-quality video using Wan2.6, with smoother motion and up to 1080P resolution. It is executed via a third-party API, so runtime depends on server-side response.",
        "io": {
          "inputs": [
            "image: Reference image to animate into a video"
          ],
          "outputs": [
            "video: Generated video from the input image"
          ]
        },
        "minComfyUIVersion": "0.3.60",
        "capabilities": {
          "workflow": [
            "image-to-video",
            "api"
          ],
          "model_options": {
            "WanImageToVideoApi": [
              "wan2.5-i2v-preview",
              "wan2.6-i2v"
            ]
          }
        }
      },
      {
        "name": "api_wan_image_to_video",
        "title": "Wan2.5: Image to Video",
        "task": "Image to Video",
        "model": "Wan2.5",
        "freshness": "established",
        "usage": 42,
        "recommend": "not_recommended",
        "description": "Transform an input image into a video with optional synchronized audio (3–30 seconds, under 15MB) using Wan2.5. The workflow accepts a reference image and an audio clip, then generates a video with synced motion and sound. This workflow calls a third-party API; execution time depends on server-side response.",
        "io": {
          "inputs": [
            "image: Reference image to animate",
            "audio: Optional audio input (3-30 seconds, under 15MB) to synchronize with the video"
          ],
          "outputs": [
            "video: Generated video with synchronized audio"
          ]
        },
        "minComfyUIVersion": "0.3.60",
        "capabilities": {
          "workflow": [
            "image-to-video",
            "api"
          ],
          "model_options": {
            "WanImageToVideoApi": [
              "wan2.5-i2v-preview",
              "wan2.6-i2v"
            ]
          }
        }
      },
      {
        "name": "video_wan2_2_14B_fun_camera",
        "title": "Wan 2.2 14B Fun Camera Control",
        "task": "Video to Video",
        "model": "Wan2.2",
        "freshness": "established",
        "usage": 20,
        "recommend": "not_recommended",
        "description": "Generate videos with camera motion controls including pan, zoom, and rotation using Wan 2.2 Fun Camera Control.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.35",
        "capabilities": {
          "workflow": [
            "video-to-video"
          ]
        }
      },
      {
        "name": "video_wan2.1_fun_camera_v1.1_14B",
        "title": "Wan 2.1 Fun Camera 14B",
        "task": "Image to Video",
        "model": "Wan2.1",
        "freshness": "established",
        "usage": 16,
        "recommend": "not_recommended",
        "description": "Generate high-quality videos with advanced camera control using the full 14B model",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.35",
        "capabilities": {
          "workflow": [
            "image-to-video"
          ]
        }
      },
      {
        "name": "video_wan2.1_fun_camera_v1.1_1.3B",
        "title": "Wan 2.1 Fun Camera 1.3B",
        "task": "Image to Video",
        "model": "Wan2.1",
        "freshness": "established",
        "usage": 10,
        "recommend": "not_recommended",
        "description": "Generate dynamic videos with cinematic camera movements using Wan 2.1 Fun Camera 1.3B model.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.35",
        "capabilities": {
          "workflow": [
            "image-to-video"
          ]
        }
      },
      {
        "name": "api_google_gemini_omni_flash_i2v",
        "title": "Gemini Omni Flash: Image to Video",
        "task": "Image to Video",
        "model": "Gemini Omni",
        "freshness": "current",
        "usage": 32,
        "recommend": "not_recommended",
        "description": "Generate a video from two images using Gemini Omni Flash, which interprets natural language prompts to control duration and aspect ratio. The workflow accepts 2 input images and produces 1 video output. Ideal for creating short brand clips, dynamic social media content, and iterative video edits through conversational prompting.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.28.3",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video"
          ],
          "model_options": {
            "GeminiVideoOmniV2": [
              "Omni Flash 1.1",
              "Omni Flash"
            ]
          }
        }
      },
      {
        "name": "api_google_gemini_omni_flash_video_edit",
        "title": "Gemini Omni Flash: Video Edit",
        "task": "Video Edit",
        "model": "Gemini Omni",
        "freshness": "current",
        "usage": 43,
        "recommend": "not_recommended",
        "description": "Edit videos with natural language using Gemini Omni Flash, transforming a single input video into one edited output based on your descriptive instructions. Specify the duration and aspect ratio in your prompt, and leverage Omni's world knowledge for intuitive edits like changing backgrounds or adding atmosphere. Ideal for quick social media remixes, cinematic scene adjustments, and iterative video refinements without technical prompts.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.28.3",
        "capabilities": {
          "workflow": [
            "api",
            "video-edit"
          ],
          "model_options": {
            "GeminiVideoOmniV2": [
              "Omni Flash 1.1",
              "Omni Flash"
            ]
          }
        }
      },
      {
        "name": "video_kandinsky5_t2v",
        "title": "Kandinsky 5.0 Video Lite Text to Video",
        "task": "Text to Video",
        "model": "Kandinsky",
        "freshness": "established",
        "usage": 14,
        "recommend": "not_recommended",
        "description": "A lightweight 2B model that generates videos from English and Russian prompts with high visual quality.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.4.0",
        "capabilities": {
          "workflow": [
            "text-to-video"
          ]
        }
      },
      {
        "name": "api_kling_omni_i2v",
        "title": "Kling O1: Image to Video",
        "task": "Character Reference",
        "model": "Kling O1",
        "freshness": "established",
        "usage": 38,
        "recommend": "not_recommended",
        "description": "Generate a dynamic video from three reference images using Kling Omni, preserving character identity through character reference. Motion and camera effects are added based on text prompts. This workflow runs via a third-party API, so response time depends on server-side load.",
        "io": {
          "inputs": [
            "image×3: Reference images including character reference for identity preservation"
          ],
          "outputs": [
            "video: Generated animated video from input images"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "api",
            "character-reference"
          ]
        }
      },
      {
        "name": "api_kling_omni_v2v",
        "title": "Kling O1: Video to Video",
        "task": "Character Reference",
        "model": "Kling O1",
        "freshness": "established",
        "usage": 13,
        "recommend": "not_recommended",
        "description": "This workflow transforms a source video using Kling O1’s Video‑to‑Video API, accepting a video, a reference image for character or style guidance, and a text prompt. It outputs a transformed video with modified content, style, or characters, leveraging the model’s support for up to 10 reference images with tags. Execution depends on server‑side response as it calls a third‑party API.",
        "io": {
          "inputs": [
            "video: Source video to transform",
            "image: Reference image for style or character guidance",
            "text: Text prompt describing the desired transformation"
          ],
          "outputs": [
            "video: Transformed output video"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "api",
            "character-reference"
          ]
        }
      },
      {
        "name": "api_pruna_p_video_t2v",
        "title": "Pruna P-Video 2: Text to Video",
        "task": "Text to Video",
        "model": "P-Video-2",
        "freshness": "recent",
        "usage": 16,
        "recommend": "not_recommended",
        "description": "Generate videos from a single text prompt with Pruna P-Video-2, producing 1-20 second clips up to 1080p at 48 fps with native audio, including ambience, sound effects, and lip-synced dialogue from an optional last-frame image. Ideal for narrative shorts, motion posters, ad B-roll, and dialogue shots.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.36.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-video"
          ]
        }
      },
      {
        "name": "api_luma_ray3_3_i2v",
        "title": "Luma Ray 3.2: Image to Video",
        "task": "Image to Video",
        "model": "Luma",
        "freshness": "current",
        "usage": 13,
        "recommend": "not_recommended",
        "description": "Use Luma Ray 3.2 to turn an input image and text prompt into a cinematic video, supporting API-based execution and iterative refinement via generation ID. This older medium model from Luma Labs produces natural motion, making it suitable for short films, marketing content, and professional video production.",
        "io": {
          "inputs": [
            "image",
            "text"
          ],
          "outputs": [
            "video"
          ]
        },
        "minComfyUIVersion": "0.26.0",
        "capabilities": {
          "workflow": [
            "image-to-video",
            "api"
          ]
        }
      },
      {
        "name": "api_kling_motion_control",
        "title": "Kling2.6: Motion Control",
        "task": "Motion Control",
        "model": "Kling 2.6",
        "freshness": "established",
        "usage": 50,
        "recommend": "low",
        "description": "Apply precise character actions and expressions from a reference video to your character image using Kling 2.6 motion control. Inputs include a character image and a reference video; the output is a result video with synchronized motion transferred from the reference. This workflow calls a third-party API, so execution time depends on server-side response.",
        "io": {
          "inputs": [
            "image: Character image to apply motion to",
            "video: Reference video providing the motion and expressions to transfer"
          ],
          "outputs": [
            "video: Result video with motion transferred from the reference"
          ]
        },
        "minComfyUIVersion": "0.7.0",
        "capabilities": {
          "workflow": [
            "motion-control",
            "api"
          ],
          "model_options": {
            "KlingMotionControl": [
              "kling-v3",
              "kling-v2-6"
            ]
          }
        }
      },
      {
        "name": "video_wan2_1_infinitetalk",
        "title": "InfiniteTalk: Audio-Driven Full-Body Video Dubbing",
        "task": "Lip Sync",
        "model": "Wan2.1 InfiniteTalk",
        "freshness": "established",
        "usage": 73,
        "recommend": "low",
        "description": "Upload a source Image and target audio. Generate a full-body dubbed video with synchronized motion while preserving the original identity, background, and camera movement.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.17.0",
        "capabilities": {
          "workflow": [
            "audio-to-video",
            "lip-sync"
          ]
        }
      },
      {
        "name": "api_kling_omni_edit_video",
        "title": "Kling O1: Video Edit",
        "task": "Video Edit",
        "model": "Kling",
        "freshness": "established",
        "usage": 38,
        "recommend": "not_recommended",
        "description": "Edit videos by combining a text prompt with a reference clip—Kling’s cinematic fluidity and style adaptation enable high-quality style transfers, element additions, and background modifications. The template outputs a generated video and supports both text-to-video and image-to-video inputs via the API.",
        "io": {
          "inputs": [
            "video: Input video to edit",
            "image×5: Reference images for style, elements, or visual guidance"
          ],
          "outputs": [
            "video: Edited video with applied style transfer or modifications"
          ]
        },
        "minComfyUIVersion": "0.14.0",
        "capabilities": {
          "workflow": [
            "api",
            "video-edit",
            "text-to-video",
            "image-to-video"
          ]
        }
      },
      {
        "name": "api_kling2_6_t2v",
        "title": "Kling2.6: Storytelling Videos with Audio",
        "task": "Text to Video",
        "model": "Kling",
        "freshness": "established",
        "usage": 27,
        "recommend": "not_recommended",
        "description": "This workflow uses Kling 2.6 to generate storytelling videos with synchronized audio from text prompts. It accepts a structured input describing the scene, subject, movement, and audio, then outputs a video with dialogue or sound effects. Execution relies on a third-party API, so response time depends on the server.",
        "io": {
          "inputs": [
            "text: Structured text prompt including scene, subject, movement, and audio description"
          ],
          "outputs": [
            "video: Generated storytelling video with synchronized dialogue or sound effects"
          ]
        },
        "minComfyUIVersion": "0.5.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-video"
          ]
        }
      },
      {
        "name": "api_kling2_6_i2v",
        "title": "Kling2.6: Animate Images with Audio",
        "task": "Image to Video",
        "model": "Kling 2.6",
        "freshness": "established",
        "usage": 13,
        "recommend": "not_recommended",
        "description": "Upload a static reference image to generate an animated video with synchronized audio using Kling 2.6. This image-to-video workflow calls a third-party API, so execution time depends on server response. Output is a video file containing the animated scene and corresponding sound effects or dialogue.",
        "io": {
          "inputs": [
            "image: Reference image to animate"
          ],
          "outputs": [
            "video: Generated animated video with synchronized audio narration or sound effects"
          ]
        },
        "minComfyUIVersion": "0.5.0",
        "capabilities": {
          "workflow": [
            "image-to-video",
            "api"
          ]
        }
      },
      {
        "name": "api_minimax_h3_flf2v",
        "title": "MiniMax H3: FLF2V",
        "task": "FLF2V",
        "model": "MiniMax H3",
        "freshness": "current",
        "usage": 64,
        "recommend": "low",
        "description": "Generate a video from a first and last frame using MiniMax H3, producing a smooth animation between the two images. You provide 2 images, and the workflow outputs a single video clip. Ideal for creating seamless transitions in animation, visualizing before-and-after transformations, and generating consistent motion for character or scene sequences.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.30.0",
        "capabilities": {
          "workflow": [
            "api",
            "flf2v"
          ],
          "model_options": {
            "MinimaxHailuo03FirstLastFrameNode": [
              "MiniMax H3"
            ]
          }
        }
      },
      {
        "name": "api_minimax_h3_t2v",
        "title": "MiniMax H3: Text to Video",
        "task": "Text to Video",
        "model": "MiniMax H3",
        "freshness": "current",
        "usage": 10,
        "recommend": "not_recommended",
        "description": "Generate video directly from text using the MiniMax H3 model, producing a 5-15 second clip with native stereo audio. Ideal for rapid concept visualization, short-form social media content, and pre-visualization for film or advertising projects.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.30.0",
        "capabilities": {
          "workflow": [
            "text-to-video",
            "api"
          ],
          "model_options": {
            "MinimaxHailuo03TextToVideoNode": [
              "MiniMax H3"
            ]
          }
        }
      },
      {
        "name": "api_minimax_h3_max_turbo_t2v",
        "title": "MiniMax H3 Turbo: Text to Video",
        "task": "Text to Video",
        "model": "MiniMax H3",
        "freshness": "recent",
        "usage": 9,
        "recommend": "not_recommended",
        "description": "Generate 5–15 second videos from text prompts using MiniMax H3 Max Turbo, a high-speed text-to-video model optimized for rapid inference. This workflow takes no file inputs and outputs a single video file, delivering 768p clips in roughly 1–3 seconds of generation time. Ideal for quick social media content, iterative creative exploration, and fast prototyping of video concepts.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.34.4",
        "capabilities": {
          "workflow": [
            "api"
          ]
        }
      },
      {
        "name": "video_capybara_v0_1_video_edit",
        "title": "Capybara: Video Edit",
        "task": "Video Edit",
        "model": "Capybara",
        "freshness": "established",
        "usage": 19,
        "recommend": "not_recommended",
        "description": "Upload an image or video and a text instruction to perform generation or editing tasks. Generate a new or edited image or video based on the provided instruction.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "video-edit"
          ]
        }
      },
      {
        "name": "video_capybara_v0_1_image_to_video",
        "title": "Capybara: Image to Video",
        "task": "Image to Video",
        "model": "Capybara",
        "freshness": "established",
        "usage": 32,
        "recommend": "not_recommended",
        "description": "Upload an image and provide a text instruction. Generate a new video where the input is edited according to your instruction, preserving structure and temporal coherence.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.3.71",
        "capabilities": {
          "workflow": [
            "image-to-video"
          ]
        }
      },
      {
        "name": "api_seedance2_5_t2v_1080p",
        "title": "Seedance2.5 1080P: Text to Video",
        "task": "Text to Video",
        "model": "Seedance 2.5",
        "freshness": "recent",
        "usage": 28,
        "recommend": "not_recommended",
        "description": "Generate cinematic 1080P video clips from text descriptions using Seedance 2.5, supporting clips up to 30 seconds with synchronized sound, music, and dialogue across any aspect ratio. Ideal for short comedy scenes, product showcases, and creative storytelling.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.31.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-video"
          ],
          "model_options": {
            "ByteDance2TextToVideoNode": [
              "Seedance 2.5",
              "Seedance 2.5 Draft",
              "Seedance 2.0",
              "Seedance 2.0 Fast",
              "Seedance 2.0 Mini"
            ]
          }
        }
      },
      {
        "name": "api_pixverse6_i2v",
        "title": "PixVerse 6.0: Image to Video",
        "task": "Image to Video",
        "model": "PixVerse",
        "freshness": "recent",
        "usage": 23,
        "recommend": "not_recommended",
        "description": "Turn a still image into a moving video clip with PixVerse V6, generating up to 15 seconds of footage at resolutions from 360p to 1080p. Supports multiple aspect ratios, integrated audio generation, and multi-shot direction for cohesive narrative flow. Ideal for social media ads, product demonstrations, and short-form storytelling.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video"
          ]
        }
      },
      {
        "name": "api_pixverse6_r2v",
        "title": "PixVerse 6.0: Reference to Video",
        "task": "Reference to Video",
        "model": "PixVerse",
        "freshness": "current",
        "usage": 10,
        "recommend": "not_recommended",
        "description": "Create reference-driven videos with PixVerse V6, generating clips that follow the style, motion, and subject traits of your uploaded reference footage. Ideal for matching a specific visual identity across variations, extending a character or scene’s look to new actions, and producing consistent branded content from a single source clip.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.34.0",
        "capabilities": {
          "workflow": [
            "api",
            "reference-to-video"
          ]
        }
      },
      {
        "name": "api_pixverse6_t2v",
        "title": "PixVerse 6.0: Text to Video",
        "task": "Text to Video",
        "model": "PixVerse",
        "freshness": "recent",
        "usage": 9,
        "recommend": "not_recommended",
        "description": "Generate videos up to 15 seconds at 1080p from text prompts with PixVerse V6, supporting multiple aspect ratios, optional integrated audio, and multi-shot scene direction in a single generation. Ideal for social media ads, product concept testing, and short-form storytelling where narrative continuity and distribution-ready framing matter.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "capabilities": {
          "workflow": [
            "api",
            "text-to-video"
          ]
        }
      },
      {
        "name": "api_pixverse_v6_max_t2v",
        "title": "PixVerse V6 Max: Text to Video",
        "task": "Text to Video",
        "model": "PixVerse",
        "freshness": "new",
        "usage": 26,
        "recommend": "medium",
        "description": "Generate videos up to 15 seconds at 1080p from a text prompt with PixVerse V6 Max, the higher-fidelity tier of PixVerse V6, with native audio and multi-shot direction in a single generation. No input files required. Ideal for social ads, product concept clips and short-form storytelling.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.42.5",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-video"
          ]
        }
      },
      {
        "name": "api_pixverse_v6_max_i2v",
        "title": "PixVerse V6 Max: Image to Video",
        "task": "Image to Video",
        "model": "PixVerse",
        "freshness": "new",
        "usage": 144,
        "recommend": "medium",
        "description": "Animate a still image with PixVerse V6 Max, keeping the aspect ratio of the input image and generating up to 15 seconds at up to 1080p with an optional native audio track. Ideal for bringing product shots, character art and keyframes to life for social and advertising use.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.42.5",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video"
          ]
        }
      },
      {
        "name": "api_pixverse_t2v",
        "title": "PixVerse: Text to Video",
        "task": "Text to Video",
        "model": "PixVerse",
        "freshness": "established",
        "usage": 8,
        "recommend": "not_recommended",
        "description": "Generate videos from text descriptions using PixVerse, with accurate prompt interpretation and stylized dynamics. This workflow takes a text prompt as input and outputs a generated video. It calls a third-party API, so execution time depends on server-side response.",
        "io": {
          "inputs": [
            "text: Text prompt describing the desired video"
          ],
          "outputs": [
            "video: Generated video based on the text prompt with PixVerse dynamics"
          ]
        },
        "minComfyUIVersion": "0.3.32",
        "capabilities": {
          "workflow": [
            "text-to-video",
            "api"
          ]
        }
      },
      {
        "name": "api_vidu_q2_flf2v",
        "title": "Vidu Q2: FLF2V",
        "task": "FLF2V",
        "model": "Vidu Q2",
        "freshness": "established",
        "usage": 21,
        "recommend": "not_recommended",
        "description": "Create smooth video transitions between start and end frames using Vidu Q2. This FLF2V workflow accepts two images and generates a transition video with cinematic camera movement and multi-subject consistency. It calls a third‑party API; execution time depends on server response.",
        "io": {
          "inputs": [
            "image×2: Start frame and end frame defining the video transition boundaries"
          ],
          "outputs": [
            "video: Generated transition video between the two frames"
          ]
        },
        "minComfyUIVersion": "0.9.0",
        "capabilities": {
          "workflow": [
            "flf2v",
            "api"
          ],
          "model_options": {
            "Vidu2StartEndToVideoNode": [
              "viduq2-pro-fast",
              "viduq2-pro",
              "viduq2-turbo"
            ]
          }
        }
      },
      {
        "name": "api_vidu_q2_r2v",
        "title": "Vidu Q2: Reference to Video",
        "task": "Character Reference",
        "model": "Vidu Q2",
        "freshness": "established",
        "usage": 26,
        "recommend": "not_recommended",
        "description": "Generate a video from two reference images using the Vidu Q2 model to preserve character identity and style continuity. This character-reference workflow supports cinematic camera movements and relies on a third-party API, so execution time depends on the server response.",
        "io": {
          "inputs": [
            "image×2: Reference images for subject identity and style consistency"
          ],
          "outputs": [
            "video: Generated video preserving reference subject identity"
          ]
        },
        "minComfyUIVersion": "0.9.0",
        "capabilities": {
          "workflow": [
            "character-reference",
            "image-to-video",
            "api"
          ],
          "model_options": {
            "Vidu2ReferenceVideoNode": [
              "viduq2"
            ]
          }
        }
      },
      {
        "name": "api_vidu_q3_text_to_video",
        "title": "Vidu Q3 Audio-Visual Video Generation",
        "task": "Text to Video",
        "model": "Vidu",
        "freshness": "established",
        "usage": 22,
        "recommend": "not_recommended",
        "description": "Generate anime-style videos from a text prompt using Vidu Q3 via a third-party API. This workflow produces a 1–16 second video with synchronized audio and automatic scene transitions, supporting viduq3-pro and viduq3-turbo model options.",
        "io": {
          "inputs": [
            "text: Text prompt describing the desired video content and style"
          ],
          "outputs": [
            "video: Generated video with synchronized audio"
          ]
        },
        "minComfyUIVersion": "0.12.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-video"
          ],
          "model_options": {
            "Vidu3TextToVideoNode": [
              "viduq3-pro",
              "viduq3-turbo"
            ]
          }
        }
      },
      {
        "name": "api_vidu_q2_i2v",
        "title": "Vidu Q2: Image to Video",
        "task": "Image to Video",
        "model": "Vidu Q2",
        "freshness": "established",
        "usage": 13,
        "recommend": "not_recommended",
        "description": "Transform a static image into a dynamic 1080p video with cinematic camera control and multi-subject consistency using the high-quality Vidu Q2 model. This image-to-video workflow runs via a third-party API and supports model variants viduq2-pro-fast, viduq2-pro, and viduq2-turbo. Accepts a reference image as input and outputs a generated video.",
        "io": {
          "inputs": [
            "image: Reference image to animate into a video"
          ],
          "outputs": [
            "video: Generated dynamic video from the input image"
          ]
        },
        "minComfyUIVersion": "0.9.0",
        "capabilities": {
          "workflow": [
            "image-to-video",
            "api"
          ],
          "model_options": {
            "Vidu2ImageToVideoNode": [
              "viduq2-pro-fast",
              "viduq2-pro",
              "viduq2-turbo"
            ]
          }
        }
      },
      {
        "name": "api_vidu_text_to_video",
        "title": "Vidu Q1: Text to Video",
        "task": "Text to Video",
        "model": "Vidu Q1",
        "freshness": "established",
        "usage": 19,
        "recommend": "not_recommended",
        "description": "Generate 1080p videos from text prompts using Vidu Q1. Provide a text description, and the workflow returns a video with adjustable movement amplitude and duration control. Execution relies on a third-party API, so response time depends on server-side processing.",
        "io": {
          "inputs": [
            "text: Text prompt describing the desired video"
          ],
          "outputs": [
            "video: Generated 1080p video from the text prompt using Vidu Q1"
          ]
        },
        "minComfyUIVersion": "0.3.51",
        "capabilities": {
          "workflow": [
            "text-to-video",
            "api"
          ],
          "model_options": {
            "ViduTextToVideoNode": [
              "viduq1"
            ]
          }
        }
      },
      {
        "name": "api_vidu_q2_t2v",
        "title": "Vidu Q2: Text to Video",
        "task": "Text to Video",
        "model": "Vidu Q2",
        "freshness": "established",
        "usage": 15,
        "recommend": "not_recommended",
        "description": "Generate high-quality video from a text prompt using the Vidu Q2 model via a third-party API. The model is known for quality and supports reference images, but this workflow uses only text. Execution time depends on server-side response.",
        "io": {
          "inputs": [
            "text: Text prompt describing the desired video content"
          ],
          "outputs": [
            "video: Generated high-quality video from the text prompt"
          ]
        },
        "minComfyUIVersion": "0.9.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-video"
          ],
          "model_options": {
            "Vidu2TextToVideoNode": [
              "viduq2"
            ]
          }
        }
      },
      {
        "name": "api_vidu_image_to_video",
        "title": "Vidu Q1: Image to Video",
        "task": "Image to Video",
        "model": "Vidu Q1",
        "freshness": "established",
        "usage": 14,
        "recommend": "not_recommended",
        "description": "Use the Vidu Q1 model to convert a static input image into a dynamic 1080p video animation. This workflow requires a single reference image and outputs a generated video. It executes as a third-party API call, so response time depends on server-side availability.",
        "io": {
          "inputs": [
            "image: Reference image to animate into a dynamic video"
          ],
          "outputs": [
            "video: Generated 1080p video animation from the input image using Vidu Q1"
          ]
        },
        "minComfyUIVersion": "0.3.51",
        "capabilities": {
          "workflow": [
            "image-to-video",
            "api"
          ],
          "model_options": {
            "ViduImageToVideoNode": [
              "viduq1"
            ]
          }
        }
      },
      {
        "name": "api_vidu_reference_to_video",
        "title": "Vidu Q1: Reference to Video",
        "task": "Image to Video",
        "model": "Vidu Q1",
        "freshness": "established",
        "usage": 11,
        "recommend": "not_recommended",
        "description": "Generate an animated video from a reference image using Vidu Q1. Provide an image and a motion or style prompt to drive the animation. The model outputs high-resolution 1080p video, and this workflow is designed for API-based execution.",
        "io": {
          "inputs": [
            "image×3: Multiple reference images showing different perspectives or styles for subject consistency"
          ],
          "outputs": [
            "video: Generated video preserving subject identity and style from reference images"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "image-to-video",
            "api"
          ],
          "model_options": {
            "ViduReferenceVideoNode": [
              "viduq1"
            ]
          }
        }
      },
      {
        "name": "api_vidu_q3_image_to_video",
        "title": "Vidu Q3 Image-to-Video with Audio",
        "task": "Image to Video",
        "model": "Vidu",
        "freshness": "established",
        "usage": 7,
        "recommend": "not_recommended",
        "description": "Use a reference image and text prompt to generate a 1–16 second anime-style video with synchronized audio, leveraging the Vidu Q3 model (viduq3-pro or turbo) for quality output. This workflow executes via a third-party API, so generation time depends on server response.",
        "io": {
          "inputs": [
            "image: Reference image to drive the video generation"
          ],
          "outputs": [
            "video: Generated video from the reference image with audio"
          ]
        },
        "minComfyUIVersion": "0.12.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video"
          ],
          "model_options": {
            "Vidu3ImageToVideoNode": [
              "viduq3-pro",
              "viduq3-turbo"
            ]
          }
        }
      },
      {
        "name": "api_happyhorse1_0_video_edit",
        "title": "HappyHorse: Video Edit",
        "task": "Video Edit",
        "model": "HappyHorse 1.0",
        "freshness": "established",
        "usage": 28,
        "recommend": "not_recommended",
        "description": "Edit a video using text instructions and optional reference images with HappyHorse 1.0’s Video Edit API. Supports style transfer and local replacement by taking a source video and up to 5 reference images. This workflow calls a third-party API; execution time depends on server-side response.",
        "io": {
          "inputs": [
            "video: Source video to edit",
            "image: Optional reference image for style transfer or replacement",
            "text: Text instructions describing the edits"
          ],
          "outputs": [
            "video: Edited output video"
          ]
        },
        "minComfyUIVersion": "0.20.0",
        "capabilities": {
          "workflow": [
            "video-edit",
            "api"
          ],
          "model_options": {
            "HappyHorseVideoEditApi": [
              "happyhorse-1.0-video-edit"
            ]
          }
        }
      },
      {
        "name": "api_happyhorse1_0_i2v",
        "title": "HappyHorse: Image to Video",
        "task": "Image to Video",
        "model": "HappyHorse 1.0",
        "freshness": "established",
        "usage": 20,
        "recommend": "not_recommended",
        "description": "This workflow uses HappyHorse 1.0 to generate a fluid video from a static image and a text prompt describing desired motion. It outputs up to 1080p video with realistic movement and intelligent semantic comprehension, relying on a third-party API for execution.",
        "io": {
          "inputs": [
            "image: Input image to animate",
            "text: Text prompt describing the desired motion"
          ],
          "outputs": [
            "video: Generated video with fluid motion"
          ]
        },
        "minComfyUIVersion": "0.20.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video"
          ],
          "model_options": {
            "HappyHorseImageToVideoApi": [
              "happyhorse-1.1-i2v",
              "happyhorse-1.0-i2v"
            ]
          }
        }
      },
      {
        "name": "api_happyhorse1_0_t2v",
        "title": "HappyHorse: Text to Video",
        "task": "Text to Video",
        "model": "HappyHorse 1.0",
        "freshness": "established",
        "usage": 17,
        "recommend": "not_recommended",
        "description": "Generate synchronized audio-video clips from text prompts using the HappyHorse 1.0 API. This workflow produces dynamic, high-quality videos with fluid motion and accurate semantic comprehension. Input a text description and receive a generated video; note that execution depends on third-party server response.",
        "io": {
          "inputs": [
            "text: Text prompt describing the video content"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.20.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-video"
          ],
          "model_options": {
            "HappyHorseTextToVideoApi": [
              "happyhorse-1.1-t2v",
              "happyhorse-1.0-t2v"
            ]
          }
        }
      },
      {
        "name": "api_happyhorse1_0_r2v",
        "title": "HappyHorse: Reference to Video",
        "task": "Reference to Video",
        "model": "HappyHorse 1.0",
        "freshness": "established",
        "usage": 8,
        "recommend": "not_recommended",
        "description": "HappyHorse 1.0 Reference-to-Video generates a short video preserving character identity from two reference images and a text scene description. Input up to 2 subject images and a prompt, output a 3–15 second video with consistent appearance. This workflow calls the HappyHorse API.",
        "io": {
          "inputs": [
            "image×2: Reference images of the subject or character",
            "text: Text prompt describing the scene"
          ],
          "outputs": [
            "video: Generated video with consistent character features"
          ]
        },
        "minComfyUIVersion": "0.20.0",
        "capabilities": {
          "workflow": [
            "reference-to-video",
            "api"
          ],
          "model_options": {
            "HappyHorseReferenceVideoApi": [
              "happyhorse-1.1-r2v",
              "happyhorse-1.0-r2v"
            ]
          }
        }
      },
      {
        "name": "ltxv_image_to_video",
        "title": "LTXV Image to Video",
        "task": "Image to Video",
        "model": "LTX-0.9.5",
        "freshness": "established",
        "usage": 15,
        "recommend": "not_recommended",
        "description": "Generate videos from still images.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.3.31",
        "capabilities": {
          "workflow": [
            "image-to-video"
          ]
        }
      },
      {
        "name": "ltxv_text_to_video",
        "title": "LTXV Text to Video",
        "task": "Text to Video",
        "model": "LTX-0.9.5",
        "freshness": "established",
        "usage": 16,
        "recommend": "not_recommended",
        "description": "This workflow generates a video from a text prompt using the LTX-0.9.5 model by Lightricks, designed for fast text-to-video generation. It takes a text description as input and outputs a short video clip, running efficiently on Comfy Cloud.",
        "io": {
          "inputs": [
            "text: Text prompt describing the video"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.3.31",
        "capabilities": {
          "workflow": [
            "text-to-video"
          ]
        }
      },
      {
        "name": "api_grok_video_edit",
        "title": "Grok: Video Edit",
        "task": "Video Edit",
        "model": "Grok",
        "freshness": "established",
        "usage": 40,
        "recommend": "not_recommended",
        "description": "Upload a video to edit, provide a text prompt describing desired modifications, and generate an edited video using the fast, high-quality Grok API. This workflow enables video editing through xAI's Grok model, combining text instructions with existing video input.",
        "io": {
          "inputs": [
            "video: Source video to edit",
            "text: Text instructions describing the modifications"
          ],
          "outputs": [
            "video: Edited output video"
          ]
        },
        "minComfyUIVersion": "0.11.1",
        "capabilities": {
          "workflow": [
            "api",
            "video-edit"
          ],
          "model_options": {
            "GrokVideoEditNode": [
              "grok-imagine-video"
            ]
          }
        }
      },
      {
        "name": "template_ltx2_3_style_transition",
        "title": "LTX-2.3 Style Transition",
        "task": "FLF2V",
        "model": "LTX-2.3",
        "freshness": "established",
        "usage": 24,
        "recommend": "not_recommended",
        "description": "Apply the ltx2.3-transition LoRA to create smooth style and scene transitions. This workflow uses a specialized model to blend visual elements.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.16.4",
        "capabilities": {
          "workflow": [
            "flf2v",
            "lora"
          ]
        }
      },
      {
        "name": "api_seedance2_0_mini_t2v",
        "title": "Seedance 2.0 Mini: Text to Video",
        "task": "Text to Video",
        "model": "Seedance 2.0",
        "freshness": "current",
        "usage": 20,
        "recommend": "not_recommended",
        "description": "Generate cinematic videos from text prompts using Dreamina’s Seedance 2.0 Mini via API. This workflow takes only a text prompt and outputs a generated video with smooth motion and consistent character generation. Supports AI camera controls and real human identity verification. Ideal for AI filmmaking, branded content, and multi-shot storytelling.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.26.2",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-video"
          ],
          "model_options": {
            "ByteDance2TextToVideoNode": [
              "Seedance 2.5",
              "Seedance 2.5 Draft",
              "Seedance 2.0",
              "Seedance 2.0 Fast",
              "Seedance 2.0 Mini"
            ]
          }
        }
      },
      {
        "name": "api_seedance2_0_t2v",
        "title": "Seedance 2.0: Text to Video",
        "task": "Text to Video",
        "model": "Seedance 2.0",
        "freshness": "established",
        "usage": 24,
        "recommend": "not_recommended",
        "description": "This workflow uses Seedance 2.0 via API to generate cinematic video from a text prompt. Provide a text description of the scene, motion, and style; the model outputs smooth, film-like motion with optional synchronized audio. You can select the Seedance 2.0 Fast variant for lower latency.",
        "io": {
          "inputs": [
            "text: Text prompt describing the desired video scene, motion, and style"
          ],
          "outputs": [
            "video: Generated video from Seedance 2.0 Text-to-Video"
          ]
        },
        "minComfyUIVersion": "0.26.2",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-video"
          ],
          "model_options": {
            "ByteDance2TextToVideoNode": [
              "Seedance 2.5",
              "Seedance 2.5 Draft",
              "Seedance 2.0",
              "Seedance 2.0 Fast",
              "Seedance 2.0 Mini"
            ]
          }
        }
      },
      {
        "name": "api_seedance2_0_r2v_real_human",
        "title": "Seedance2.0 Real Human: Reference to Video",
        "task": "Reference to Video",
        "model": "Seedance 2.0",
        "freshness": "established",
        "usage": 12,
        "recommend": "not_recommended",
        "description": "Upload a reference video of a real person for identity verification. Use your own photo to validate and generate a personalized video output.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.26.2",
        "capabilities": {
          "workflow": [
            "api",
            "reference-to-video"
          ],
          "model_options": {
            "ByteDance2ReferenceNodeV2": [
              "Seedance 2.5",
              "Seedance 2.5 Draft",
              "Seedance 2.0",
              "Seedance 2.0 Fast",
              "Seedance 2.0 Mini"
            ]
          }
        }
      },
      {
        "name": "template_seedance_2_0_plus_llm_prompt_helper",
        "title": "Seedance 2.0 + LLM Prompt Helper",
        "task": "Reference to Video",
        "model": "Gemini-3.1-Pro",
        "freshness": "current",
        "usage": 10,
        "recommend": "not_recommended",
        "description": "Generate cinematic videos from two reference images and optional text. Gemini 3.1 Pro automatically creates a detailed Seedance 2.0-style prompt, executed via Seedance 2.0 API for reference-to-video output. Third-party API execution means response times depend on server-side processing.",
        "io": {
          "inputs": [
            "image×2: Reference images for video generation (e.g. subject, style, location)",
            "text: Optional text description to guide LLM prompt generation"
          ],
          "outputs": [
            "video: Generated video from Seedance 2.0"
          ]
        },
        "minComfyUIVersion": "0.35.2",
        "capabilities": {
          "workflow": [
            "api",
            "reference-to-video"
          ],
          "model_options": {
            "GeminiNode": [
              "gemini-2.5-pro",
              "gemini-2.5-flash",
              "gemini-3-pro-preview",
              "gemini-3-1-pro",
              "gemini-3-1-flash-lite"
            ]
          }
        }
      },
      {
        "name": "api_seedance2_0_flf2v_real_human",
        "title": "Seedance2.0 Real Human: FLF2V",
        "task": "FLF2V",
        "model": "Seedance 2.0",
        "freshness": "established",
        "usage": 9,
        "recommend": "not_recommended",
        "description": "Generate a cinematic, identity-consistent video by providing two images (first and last frames) of a real person. Seedance 2.0 interpolates between them using first-last-frame-to-video with real-human identity verification. This workflow runs via a third-party API for server-side inference.",
        "io": {
          "inputs": [
            "image×2: First and last frames of a real person for identity-verified video generation"
          ],
          "outputs": [
            "video: Generated video of a verified real person from first and last frame control"
          ]
        },
        "minComfyUIVersion": "0.26.2",
        "capabilities": {
          "workflow": [
            "api",
            "flf2v"
          ],
          "model_options": {
            "ByteDance2FirstLastFrameNode": [
              "Seedance 2.5",
              "Seedance 2.5 Draft",
              "Seedance 2.0",
              "Seedance 2.0 Fast",
              "Seedance 2.0 Mini"
            ]
          }
        }
      },
      {
        "name": "video_humo",
        "title": "HuMo Video Generation",
        "task": "Text to Video",
        "model": "HuMo",
        "freshness": "established",
        "usage": 17,
        "recommend": "not_recommended",
        "description": "Generate videos based on audio, image, and text, keeping the character's lip sync.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.60",
        "capabilities": {
          "workflow": [
            "text-to-video"
          ]
        }
      },
      {
        "name": "template_sirolim_image_script_video",
        "title": "Image to Script to Kling Video Generation",
        "task": "Image to Video",
        "model": "Google Gemini",
        "freshness": "established",
        "usage": 18,
        "recommend": "not_recommended",
        "description": "Use Google Gemini to convert a text prompt into a descriptive script, then send that script to the Kling 3.0 Omni API for video generation. The workflow outputs the resulting video, showcasing Gemini's strong script understanding and third‑party API integration.",
        "io": {
          "inputs": [
            "image: Reference image to base the video on",
            "text: Brief prompt with important details and output aspect ratio"
          ],
          "outputs": [
            "video: Generated video from the image and script"
          ]
        },
        "minComfyUIVersion": "0.35.2",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video"
          ]
        }
      },
      {
        "name": "api_kling_o3_flf2v",
        "title": "Kling O3: FLF2V",
        "task": "FLF2V",
        "model": "Kling O3",
        "freshness": "established",
        "usage": 10,
        "recommend": "not_recommended",
        "description": "Generate videos with precise semantic control by providing start and end frame images using Kling O3's First-Last Frame to Video (FLF2V) API. This workflow supports longer durations and improved narrative coherence through multiple reference images. It calls a third-party API, so execution time depends on server-side response.",
        "io": {
          "inputs": [
            "image×4: Start frame and end frame images defining the video boundaries"
          ],
          "outputs": [
            "video: Generated video sequence"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "api",
            "flf2v"
          ]
        }
      },
      {
        "name": "api_kling_o3_t2v",
        "title": "Kling O3: Text to Video",
        "task": "Text to Video",
        "model": "Kling O3",
        "freshness": "established",
        "usage": 25,
        "recommend": "not_recommended",
        "description": "This ComfyUI workflow uses the Kling O3 API to generate 15‑second videos from text prompts. Input a text description and receive a video with synchronized audio and character consistency. It is a third‑party API workflow, so execution time depends on server‑side response.",
        "io": {
          "inputs": [
            "text: Text prompt describing the video content"
          ],
          "outputs": [
            "video: Generated video with synchronized audio"
          ]
        },
        "minComfyUIVersion": "0.3.76",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-video"
          ]
        }
      },
      {
        "name": "api_ltx2_5_t2v",
        "title": "LTX-2.5 (Pro): Text to Video",
        "task": "Text to Video",
        "model": "LTX-2.5",
        "freshness": "recent",
        "usage": 12,
        "recommend": "not_recommended",
        "description": "Generate video from text with LTX 2.5, a video model built for fast, high-quality production work. Type a prompt and get a cinematic clip with sharp detail, accurate motion, and support for native multishot scenes that hold character and style across cuts.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.32.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-video"
          ]
        }
      },
      {
        "name": "video_kandinsky5_i2v",
        "title": "Kandinsky 5.0 Video Lite Image to Video",
        "task": "Image to Video",
        "model": "Kandinsky",
        "freshness": "established",
        "usage": 14,
        "recommend": "not_recommended",
        "description": "A lightweight 2B model that generates videos from English and Russian prompts with high visual quality.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.4.0",
        "capabilities": {
          "workflow": [
            "image-to-video"
          ]
        }
      },
      {
        "name": "api_wan_text_to_video",
        "title": "Wan2.5: Text to Video",
        "task": "Image to Video",
        "model": "Wan2.5",
        "freshness": "established",
        "usage": 16,
        "recommend": "not_recommended",
        "description": "This workflow generates a video from a text prompt using Wan2.5 model variants (wan2.5-t2v-preview, wan2.6-t2v). Optionally, you can sync an audio file (3–30 seconds, under 15MB) to enhance motion quality. The output is a video with synchronized audio. This workflow uses a third‑party API, so execution time depends on server‑side response.",
        "io": {
          "inputs": [
            "text: Text prompt describing the desired video content",
            "audio: Optional audio input to synchronize with generated video (3-30 seconds, under 15MB)"
          ],
          "outputs": [
            "video: Generated video with synchronized audio and enhanced motion quality"
          ]
        },
        "minComfyUIVersion": "0.3.60",
        "capabilities": {
          "workflow": [
            "image-to-video",
            "api"
          ],
          "model_options": {
            "WanTextToVideoApi": [
              "wan2.5-t2v-preview",
              "wan2.6-t2v"
            ]
          }
        }
      },
      {
        "name": "video_wan2_2_5B_fun_control",
        "title": "Wan 2.2 5B Fun Control",
        "task": "Text to Video",
        "model": "Wan2.2",
        "freshness": "established",
        "usage": 14,
        "recommend": "not_recommended",
        "description": "Multi-condition video control with pose, depth, and edge guidance. Compact 5B size for experimental development.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.50",
        "capabilities": {
          "workflow": [
            "text-to-video",
            "controlnet"
          ]
        }
      },
      {
        "name": "api_wan2_6_t2v",
        "title": "Wan2.6: Text to Video",
        "task": "Text to Video",
        "model": "Wan2.6",
        "freshness": "established",
        "usage": 11,
        "recommend": "not_recommended",
        "description": "Generate high-quality videos from text prompts using the Wan2.6 model, with support for 1080P resolution and smoother motion. This text-to-video workflow calls a third-party API—execution depends on server response—and accepts a single text prompt describing the desired scene, outputting a generated video.",
        "io": {
          "inputs": [
            "text: Text prompt describing the desired video scene and motion"
          ],
          "outputs": [
            "video: Generated video matching the text description"
          ]
        },
        "minComfyUIVersion": "0.3.60",
        "capabilities": {
          "workflow": [
            "text-to-video",
            "api"
          ],
          "model_options": {
            "WanTextToVideoApi": [
              "wan2.5-t2v-preview",
              "wan2.6-t2v"
            ]
          }
        }
      },
      {
        "name": "wan2.1_flf2v_720_f16",
        "title": "Wan 2.1 FLF2V 720p F16",
        "task": "FLF2V",
        "model": "Wan2.1",
        "freshness": "established",
        "usage": 9,
        "recommend": "not_recommended",
        "description": "Generate videos by controlling first and last frames using Wan 2.1 FLF2V.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.31",
        "capabilities": {
          "workflow": [
            "flf2v"
          ]
        }
      },
      {
        "name": "api_wan_r2v",
        "title": "WAN2.6: Reference to Video",
        "task": "Character Reference",
        "model": "Wan2.6",
        "freshness": "established",
        "usage": 10,
        "recommend": "not_recommended",
        "description": "This identity-preserving video workflow uses two reference videos—one for identity and one for motion guidance—to generate cinematic output with natural movement. It relies on the Wan2.6 model and executes via a third-party API, with response time depending on server load.",
        "io": {
          "inputs": [
            "video×2: Reference videos for identity and motion guidance"
          ],
          "outputs": [
            "video: Generated video preserving subject identity from reference"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "api",
            "character-reference"
          ],
          "model_options": {
            "WanReferenceVideoApi": [
              "wan2.6-r2v"
            ]
          }
        }
      },
      {
        "name": "video_wan_vace_14B_t2v",
        "title": "Wan2.1 VACE Text to Video",
        "task": "Text to Video",
        "model": "Wan2.1",
        "freshness": "established",
        "usage": 10,
        "recommend": "not_recommended",
        "description": "Transform text descriptions into high-quality videos. Supports both 480p and 720p with VACE-14B model.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.3.31",
        "capabilities": {
          "workflow": [
            "text-to-video"
          ]
        }
      },
      {
        "name": "video_wanmove_480p",
        "title": "Wan-Move Motion-Control Image to Video",
        "task": "Image to Video",
        "model": "Wan",
        "freshness": "established",
        "usage": 8,
        "recommend": "not_recommended",
        "description": "Generate videos from a single image using Wan-Move, with fine-grained point-level motion control via trajectory guidance.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.5.0",
        "capabilities": {
          "workflow": [
            "image-to-video",
            "motion-control"
          ]
        }
      },
      {
        "name": "video_wan_ati",
        "title": "Wan2.1 ATI",
        "task": "Motion Control",
        "model": "Wan2.1",
        "freshness": "established",
        "usage": 9,
        "recommend": "not_recommended",
        "description": "Trajectory-controlled Video Generation.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.46",
        "capabilities": {
          "workflow": [
            "motion-control"
          ]
        }
      },
      {
        "name": "wan2.1_fun_inp",
        "title": "Wan 2.1 Inpainting",
        "task": "Video Generation",
        "model": "Wan2.1",
        "freshness": "established",
        "usage": 9,
        "recommend": "not_recommended",
        "description": "This Wan 2.1 inpainting workflow edits an input image based on text instructions, using a legacy video generation model optimized for fast, low-quality output with video Lora support. It takes an image and text prompt, then returns the edited image via inpainting.",
        "io": {
          "inputs": [
            "image: Input image to edit",
            "text: Edit instructions"
          ],
          "outputs": [
            "image: Edited image"
          ]
        },
        "minComfyUIVersion": "0.3.31",
        "capabilities": {
          "workflow": [
            "inpainting"
          ]
        }
      },
      {
        "name": "video_wan2.1_alpha_t2v_14B",
        "title": "Wan2.1 Alpha T2V",
        "task": "Text to Video",
        "model": "Wan2.1",
        "freshness": "established",
        "usage": 18,
        "recommend": "not_recommended",
        "description": "Generate text-to-video with alpha channel support for transparent backgrounds and semi-transparent objects.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.3.8",
        "capabilities": {
          "workflow": [
            "text-to-video"
          ]
        }
      },
      {
        "name": "api_bfl_flux3_t2v",
        "title": "FLUX 3 Video: Text to Video",
        "task": "Text to Video",
        "model": "FLUX 3 Video",
        "freshness": "current",
        "usage": 27,
        "recommend": "not_recommended",
        "description": "Generate full videos from a single text prompt with FLUX 3, which expands natural language into complete scenes with synchronized audio, no reference images or complex settings needed. Outputs 1 video in 720p at 24fps, 5 to 20 seconds long, supporting any aspect ratio from 9:16 to 21:9 with built-in sound design. Ideal for quick social media clips, ad mockups, and concept films requiring realistic motion and cinematic style.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.30.2",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-video"
          ]
        }
      },
      {
        "name": "api_runway_aleph2_video_edit",
        "title": "Runway Aleph2: Video Edit",
        "task": "Video Edit",
        "model": "Runway",
        "freshness": "current",
        "usage": 60,
        "recommend": "low",
        "description": "This template uses Runway's frontier video model via API to perform video editing from a text prompt. It takes a text description of desired changes and outputs an edited video that preserves original motion and timing. Designed for relighting and consistent lighting across photo sets, this API-based workflow relies on third-party execution.",
        "io": {
          "inputs": [
            "text"
          ],
          "outputs": [
            "image"
          ]
        },
        "minComfyUIVersion": "0.25.0",
        "capabilities": {
          "workflow": [
            "api",
            "video-edit"
          ]
        }
      },
      {
        "name": "api_runway_gen4_turbo_image_to_video",
        "title": "Runway: Gen4 Turbo Image to Video",
        "task": "Image to Video",
        "model": "Runway",
        "freshness": "established",
        "usage": 26,
        "recommend": "not_recommended",
        "description": "Generate dynamic videos from images using Runway Gen4 Turbo.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.3.37",
        "capabilities": {
          "workflow": [
            "image-to-video",
            "api"
          ]
        }
      },
      {
        "name": "api_heygen_video_translate",
        "title": "HeyGen: Video Translate",
        "task": "Lip Sync",
        "model": "HeyGen",
        "freshness": "current",
        "usage": 11,
        "recommend": "not_recommended",
        "description": "Translate video content into another language with realistic lip-sync via HeyGen's API. It transcribes speech, translates text, generates a cloned voice, and re-syncs mouth movements. Takes an input video and a target language prompt, outputs a fully dubbed video. Ideal for localizing talking-head presentations.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.28.1",
        "capabilities": {
          "workflow": [
            "api",
            "lip-sync"
          ]
        }
      },
      {
        "name": "api_heygen_avatar_video",
        "title": "HeyGen: Avatar Video",
        "task": "Lip Sync",
        "model": "HeyGen",
        "freshness": "current",
        "usage": 11,
        "recommend": "not_recommended",
        "description": "Generate a talking‑head avatar video from a text prompt using the HeyGen API, with realistic lip‑sync animation. This API‑only workflow takes a text input and outputs a short clip suitable for digital presenters or social media content.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.28.1",
        "capabilities": {
          "workflow": [
            "api",
            "lip-sync"
          ]
        }
      },
      {
        "name": "video_ltx2_i2v_lora",
        "title": "Squish Anything with LTX-2 I2V and LoRA",
        "task": "Image to Video",
        "model": "LTX-2",
        "freshness": "established",
        "usage": 11,
        "recommend": "not_recommended",
        "description": "Upload any image to apply the Squish It LoRA effect. Generate a squished animation using the LTX-2 I2V model.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "image-to-video"
          ]
        }
      },
      {
        "name": "video_ltx2_t2v_distilled",
        "title": "LTX-2 Text to Video (Distilled)",
        "task": "Text to Video",
        "model": "LTX-2",
        "freshness": "established",
        "usage": 17,
        "recommend": "not_recommended",
        "description": "Generate high-quality videos from text prompts with synchronized audio-video generation using LTX-2 distilled model. Optimized for faster generation while maintaining quality. Features expressive lip sync, dynamic motion generation, and improved speed.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "text-to-video"
          ]
        }
      },
      {
        "name": "video_ltx2_depth_to_video",
        "title": "LTX-2 Depth to Video",
        "task": "Depth Estimation",
        "model": "LTX-2",
        "freshness": "established",
        "usage": 11,
        "recommend": "not_recommended",
        "description": "Generate high-quality videos from depth maps with synchronized audio-video generation using LTX-2. Features precise depth control, spatial awareness, and efficient performance.",
        "io": {
          "inputs": [
            "image: Input image"
          ],
          "outputs": [
            "image: Depth map"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "controlnet"
          ]
        }
      },
      {
        "name": "api_happyhorse1_1_i2v",
        "title": "HappyHorse 1.1: Image to Video",
        "task": "Image to Video",
        "model": "HappyHorse 1.1",
        "freshness": "current",
        "usage": 11,
        "recommend": "not_recommended",
        "description": "Generate a short video from a static first frame with synchronized audio using HappyHorse 1.1. This image-to-video workflow accepts one reference image and a motion prompt via API, outputting an animated clip ideal for ads or product demos.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.26.1",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video"
          ],
          "model_options": {
            "HappyHorseImageToVideoApi": [
              "happyhorse-1.1-i2v",
              "happyhorse-1.0-i2v"
            ]
          }
        }
      },
      {
        "name": "api_happyhorse1_1_r2v",
        "title": "HappyHorse 1.1: Reference to Video",
        "task": "Reference to Video",
        "model": "HappyHorse 1.1",
        "freshness": "current",
        "usage": 9,
        "recommend": "not_recommended",
        "description": "Generate video from reference images using HappyHorse 1.1, locking character and scene appearance while following a text prompt for motion. This API workflow takes an image reference and a text prompt, producing an animated video with synchronized audio. Ideal for short films, ads, and game cutscenes requiring consistent character design.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.26.1",
        "capabilities": {
          "workflow": [
            "api",
            "reference-to-video"
          ],
          "model_options": {
            "HappyHorseReferenceVideoApi": [
              "happyhorse-1.1-r2v",
              "happyhorse-1.0-r2v"
            ]
          }
        }
      },
      {
        "name": "api_happyhorse1_1_t2v",
        "title": "HappyHorse 1.1: Text to Video",
        "task": "Text to Video",
        "model": "HappyHorse 1.1",
        "freshness": "current",
        "usage": 3,
        "recommend": "not_recommended",
        "description": "This text-to-video workflow uses HappyHorse 1.1 to create short videos with synchronized dialogue, sound effects, and music directly from a text prompt. It supports API execution and is suited for ads, episodic shorts, and game cutscenes, maintaining character and scene consistency.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.26.1",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-video"
          ],
          "model_options": {
            "HappyHorseTextToVideoApi": [
              "happyhorse-1.1-t2v",
              "happyhorse-1.0-t2v"
            ]
          }
        }
      },
      {
        "name": "api_sync_so_talking_image",
        "title": "Sync 3: Talking Image",
        "task": "Lip Sync",
        "model": "Sync 3",
        "freshness": "current",
        "usage": 12,
        "recommend": "not_recommended",
        "description": "Generate a talking-head video from a single face photo and an audio clip, with lips and lower-face motion perfectly synchronized to the speech. Each image-audio pair produces one video while keeping the background and any objects unchanged. Ideal for product demos, e-commerce introductions, and marketing videos.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.28.0",
        "capabilities": {
          "workflow": [
            "api",
            "lip-sync"
          ],
          "model_options": {
            "SyncTalkingImageNode": [
              "sync-3"
            ]
          }
        }
      },
      {
        "name": "templates_ohneis_i2v",
        "title": "Image to Video",
        "task": "Image to Video",
        "model": "Kling 2.6",
        "freshness": "established",
        "usage": 6,
        "recommend": "not_recommended",
        "description": "This workflow takes a text prompt, generates a cinematic-style image using Gemini models, enhances it with Nano Banana Pro, and produces a video via Kling 2.6 Image-to-Video. It requires only a text input and outputs a stylized video clip. Execution relies on a third-party API, so response times vary.",
        "io": {
          "inputs": [
            "text: Text prompt describing the desired cinematic image style"
          ],
          "outputs": [
            "video: Generated video from Kling with cinematic style"
          ]
        },
        "minComfyUIVersion": "0.5.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-video",
            "text-to-image"
          ],
          "model_options": {
            "GeminiImage2Node": [
              "gemini-3-pro-image-preview",
              "Nano Banana 2 (Gemini 3.1 Flash Image)"
            ]
          }
        }
      },
      {
        "name": "api_luma_ray3_2_video_edit",
        "title": "Luma Ray 3.2: Video Edit",
        "task": "Video Edit",
        "model": "Luma",
        "freshness": "current",
        "usage": 9,
        "recommend": "not_recommended",
        "description": "Edit videos with natural language commands using the Luma Ray 3.2 model via API. This workflow takes a text description of the desired edit and processes one input video to generate a single edited image output.",
        "io": {
          "inputs": [
            "text"
          ],
          "outputs": [
            "image"
          ]
        },
        "minComfyUIVersion": "0.26.0",
        "capabilities": {
          "workflow": [
            "api",
            "video-edit"
          ]
        }
      },
      {
        "name": "api_luma_ray3_3_t2v",
        "title": "Luma Ray 3.2: Text to Video",
        "task": "Text to Video",
        "model": "Luma",
        "freshness": "current",
        "usage": 5,
        "recommend": "not_recommended",
        "description": "Generate videos from text prompts using the Luma Ray 3.2 API. This legacy Luma Labs model produces clips with natural motion. The workflow accepts a text description and outputs a short video, suitable for storyboards or quick visualizations.",
        "io": {
          "inputs": [
            "text"
          ],
          "outputs": [
            "video"
          ]
        },
        "minComfyUIVersion": "0.26.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-video"
          ]
        }
      },
      {
        "name": "api_bytedance_flf2v",
        "title": "Seedance1.0 Pro: Start End to Video",
        "task": "FLF2V",
        "model": "Seedance1.0 Pro",
        "freshness": "established",
        "usage": 13,
        "recommend": "not_recommended",
        "description": "Generate a cinematic video transition between a start frame and end frame using ByteDance's Seedance1.0 Pro model via FLF2V task. The workflow accepts two input images and outputs a smooth, professionally styled video bridging them with fluid motion and scene consistency. Designed for third-party API execution in ComfyUI.",
        "io": {
          "inputs": [
            "image×2: Start frame and end frame for the video transition"
          ],
          "outputs": [
            "video: Generated cinematic transition video between the two frames"
          ]
        },
        "minComfyUIVersion": "0.3.58",
        "capabilities": {
          "workflow": [
            "api",
            "flf2v"
          ],
          "model_options": {
            "ByteDanceFirstLastFrameNode": [
              "seedance-1-5-pro-251215",
              "seedance-1-0-pro-250528"
            ]
          }
        }
      },
      {
        "name": "api_bytedance_text_to_video",
        "title": "Seedance1.0 Pro: Text to Video",
        "task": "Text to Video",
        "model": "Seedance1.0 Pro",
        "freshness": "established",
        "usage": 7,
        "recommend": "not_recommended",
        "description": "Generate a video directly from a text prompt using ByteDance's Seedance1.0 Pro model. The workflow accepts a text description as input and outputs a generated video with natural motion and cinematic quality. Because this template calls a third-party API, execution time depends on server-side response.",
        "io": {
          "inputs": [
            "text: Text prompt describing the desired video content"
          ],
          "outputs": [
            "video: Generated video from the text prompt using Seedance1.0 Pro"
          ]
        },
        "minComfyUIVersion": "0.3.58",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-video"
          ],
          "model_options": {
            "ByteDanceTextToVideoNode": [
              "seedance-1-5-pro-251215",
              "seedance-1-0-pro-250528",
              "seedance-1-0-pro-fast-251015"
            ]
          }
        }
      },
      {
        "name": "api_wan2_7_r2v",
        "title": "Wan2.7: Reference to Video",
        "task": "Reference to Video",
        "model": "Wan2.7",
        "freshness": "established",
        "usage": 18,
        "recommend": "not_recommended",
        "description": "Generate a video with consistent character features using the Wan2.7 Reference-to-Video API. Upload two character reference images and a text prompt describing the scene; the model outputs a video that preserves the subject’s appearance. This workflow calls a third-party API, so execution time depends on server-side response.",
        "io": {
          "inputs": [
            "image×2: Character reference images for identity consistency",
            "text: Text prompt describing the scene"
          ],
          "outputs": [
            "video: Generated video with consistent character features"
          ]
        },
        "minComfyUIVersion": "0.18.4",
        "capabilities": {
          "workflow": [
            "api",
            "reference-to-video"
          ],
          "model_options": {
            "Wan2ReferenceVideoApi": [
              "wan2.7-r2v"
            ]
          }
        }
      },
      {
        "name": "api_wan2_7_t2v",
        "title": "Wan2.7: Text to Video",
        "task": "Lip Sync",
        "model": "Wan2.7",
        "freshness": "established",
        "usage": 23,
        "recommend": "not_recommended",
        "description": "Generate medium-quality videos from text prompts using the Wan2.7 model from Alibaba. This workflow optionally accepts an audio reference to create lip-synced video via API. Input: text prompt; output: generated video.",
        "io": {
          "inputs": [
            "text: Text prompt describing the video content"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.18.4",
        "capabilities": {
          "workflow": [
            "api",
            "lip-sync",
            "text-to-video"
          ],
          "model_options": {
            "Wan2TextToVideoApi": [
              "wan2.7-t2v"
            ]
          }
        }
      },
      {
        "name": "txt_to_image_to_video",
        "title": "SVD Text to Image to Video",
        "task": "Text to Video",
        "model": "SVD",
        "freshness": "established",
        "usage": 9,
        "recommend": "not_recommended",
        "description": "This text-to-video workflow generates an image from a text prompt, then animates it into a video using Stable Video Diffusion. It takes a single text description as input and outputs a synthesized video clip.",
        "io": {
          "inputs": [
            "text: Text prompt describing the video"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.3.31",
        "capabilities": {
          "workflow": [
            "text-to-video"
          ]
        }
      },
      {
        "name": "api_google_gemini_omni_flash_t2v",
        "title": "Gemini Omni Flash: Text to Video",
        "task": "Text to Video",
        "model": "Gemini Omni",
        "freshness": "current",
        "usage": 16,
        "recommend": "not_recommended",
        "description": "Generate cinematic video from natural language prompts using Gemini Omni Flash, transforming text descriptions into a single video output with world-aware motion, lighting, and sound. Ideal for social media content creation, rapid video prototyping, and iterative visual storytelling with conversational editing.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.28.3",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-video"
          ],
          "model_options": {
            "GeminiVideoOmniV2": [
              "Omni Flash 1.1",
              "Omni Flash"
            ]
          }
        }
      },
      {
        "name": "hunyuan_video_text_to_video",
        "title": "Hunyuan Video Text to Video",
        "task": "Text to Video",
        "model": "Hunyuan Video",
        "freshness": "established",
        "usage": 10,
        "recommend": "not_recommended",
        "description": "Generate videos from text prompts using Tencent's Hunyuan Video model. This workflow accepts a text description as input and outputs a generated video clip. It is designed for Comfy Cloud and executes quickly without requiring third-party API calls.",
        "io": {
          "inputs": [
            "text: Text prompt describing the video"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.3.31",
        "capabilities": {
          "workflow": [
            "text-to-video"
          ]
        }
      },
      {
        "name": "api_pixverse_i2v",
        "title": "PixVerse: Image to Video",
        "task": "Image to Video",
        "model": "PixVerse",
        "freshness": "established",
        "usage": 6,
        "recommend": "not_recommended",
        "description": "Generate dynamic videos from static images with motion and effects using PixVerse. The workflow takes a reference image as input and returns a generated video, with optional template selection. This workflow calls a third-party API, so execution time depends on the server-side response.",
        "io": {
          "inputs": [
            "image: Reference image to animate into a video"
          ],
          "outputs": [
            "video: Generated video from the input image using PixVerse"
          ]
        },
        "minComfyUIVersion": "0.3.32",
        "capabilities": {
          "workflow": [
            "image-to-video",
            "api"
          ]
        }
      },
      {
        "name": "api_pixverse_template_i2v",
        "title": "PixVerse Templates: Image to Video",
        "task": "Image to Video",
        "model": "PixVerse",
        "freshness": "established",
        "usage": 6,
        "recommend": "not_recommended",
        "description": "Generate dynamic videos from a static image using PixVerse's image-to-video template with predefined style options. The workflow accepts a reference image and returns a motion video. This template calls a third-party API, so execution time depends on server-side response.",
        "io": {
          "inputs": [
            "image: Reference image to animate into a video using a PixVerse template"
          ],
          "outputs": [
            "video: Generated video from the input image using PixVerse template"
          ]
        },
        "minComfyUIVersion": "0.3.32",
        "capabilities": {
          "workflow": [
            "image-to-video",
            "api"
          ]
        }
      },
      {
        "name": "api_kling_omni_t2v",
        "title": "Kling O1: Text to Video",
        "task": "Text to Video",
        "model": "Kling O1",
        "freshness": "established",
        "usage": 6,
        "recommend": "not_recommended",
        "description": "Kling O1, an image model with support for up to 10 reference images, generates videos from text descriptions. This workflow calls a third-party API — execution time depends on server-side response. Input a text prompt describing the desired video content and mood, and receive a generated video.",
        "io": {
          "inputs": [
            "text: Text prompt describing the desired video content and mood"
          ],
          "outputs": [
            "video: Generated video matching the text description"
          ]
        },
        "minComfyUIVersion": "0.3.76",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-video"
          ]
        }
      },
      {
        "name": "video_minimax_h3_i2v_continuation",
        "title": "Image to Video",
        "task": "Image to Video",
        "model": "MiniMax H3",
        "freshness": "recent",
        "usage": 398,
        "recommend": "medium",
        "description": "Start with an image and an empty prompt, then describe the motion, shots, and audio you want MiniMax H3 to generate. The workflow produces one video with synchronized audio and remains fully editable before execution.",
        "io": {
          "inputs": [
            "image: Reference image",
            "text: Motion or style prompt"
          ],
          "outputs": [
            "video: Animated video"
          ]
        },
        "minComfyUIVersion": "0.30.0",
        "capabilities": {
          "workflow": [
            "image-to-video",
            "int8"
          ]
        }
      },
      {
        "name": "templates_mjm_looped_restyler",
        "title": "Looped Restyler",
        "task": "FLF2V",
        "model": "Google Gemini",
        "freshness": "established",
        "usage": 6,
        "recommend": "not_recommended",
        "description": "Upload a starting image and its prompt to generate a looping video from the first to the last frame.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.53",
        "capabilities": {
          "workflow": [
            "api",
            "flf2v",
            "image-to-video"
          ]
        }
      },
      {
        "name": "video_wan2_2_5B_fun_inpaint",
        "title": "Wan 2.2 5B Fun Inpaint",
        "task": "Text to Video",
        "model": "Wan2.2",
        "freshness": "established",
        "usage": 2,
        "recommend": "not_recommended",
        "description": "Efficient video inpainting from start and end frames. 5B model delivers quick iterations for testing workflows.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.31",
        "capabilities": {
          "workflow": [
            "text-to-video"
          ]
        }
      },
      {
        "name": "video_wan2_2_14B_fun_inpaint",
        "title": "Wan 2.2 14B Fun Inpaint",
        "task": "FLF2V",
        "model": "Wan2.2",
        "freshness": "established",
        "usage": 14,
        "recommend": "not_recommended",
        "description": "Generate videos from start and end frames using Wan 2.2 Fun Inpaint.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.31",
        "capabilities": {
          "workflow": [
            "flf2v"
          ]
        }
      },
      {
        "name": "api_vidu_start_end_to_video",
        "title": "Vidu Q1: Start End to Video",
        "task": "FLF2V",
        "model": "Vidu Q1",
        "freshness": "established",
        "usage": 2,
        "recommend": "not_recommended",
        "description": "This workflow uses the Vidu Q1 model to generate a smooth video transition between a provided start frame and an end frame. It accepts two images as input and outputs a video with natural motion interpolation and consistent visual quality. The workflow executes via a third-party API, so response time depends on server-side processing.",
        "io": {
          "inputs": [
            "image×2: Start frame and end frame images to define the video transition"
          ],
          "outputs": [
            "video: Generated video with smooth interpolation between start and end frames"
          ]
        },
        "minComfyUIVersion": "0.3.51",
        "capabilities": {
          "workflow": [
            "api",
            "flf2v"
          ],
          "model_options": {
            "ViduStartEndToVideoNode": [
              "viduq1"
            ]
          }
        }
      }
    ]
  },
  {
    "category": "Video Tools",
    "description": "Tooling workflows for video processing helpers such as upscaling, frame interpolation, background removal, video extend, pose and depth extraction, and clip editing.",
    "templates": [
      {
        "name": "api_pika_h3_flex_t2v",
        "title": "Pika H3 Flex: Text to Video (API)",
        "task": "Text to Video",
        "model": "MiniMax H3",
        "freshness": "new",
        "usage": 0,
        "recommend": "medium",
        "description": "Generate a clip with synchronized audio from a text prompt with Pika H3 Flex through the partner API, no weights to download. One node covers text to video, image to video and first-last-frame by connecting the optional frame inputs, with 480p to 4K output and 5 to 15 second durations. Ideal for prompt exploration on MiniMax H3 when a local install is not available.",
        "io": {
          "inputs": [
            "text: Video prompt"
          ],
          "outputs": [
            "video: Generated video"
          ]
        },
        "minComfyUIVersion": "0.44.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-video"
          ]
        }
      },
      {
        "name": "api_pruna_p_video_upscale",
        "title": "Video Upscale: Pruna P-Video",
        "task": "Video Upscale",
        "model": "Pruna P-Video",
        "freshness": "new",
        "usage": 0,
        "recommend": "medium",
        "description": "Upscale and restore a clip with Pruna P-Video in a single node. Pick a target pixel area from 720p to 4K, optionally raise the frame rate, and let one pass rebuild detail instead of only resampling pixels. Clips already larger than the target are enhanced at their own size. Ideal for finishing low-resolution generations for delivery or social crops.",
        "io": {
          "inputs": [
            "video: Input video"
          ],
          "outputs": [
            "video: Upscaled video"
          ]
        },
        "minComfyUIVersion": "0.44.0",
        "capabilities": {
          "workflow": [
            "api",
            "video-upscale"
          ]
        }
      },
      {
        "name": "api_minimax_h3_max_camera_control",
        "title": "Camera Control: MiniMax H3 Max",
        "task": "Video Generation",
        "model": "MiniMax H3",
        "freshness": "new",
        "usage": 0,
        "recommend": "medium",
        "description": "Orbit the camera around a still image with MiniMax H3 Max, then put your own music under it. The image is the first frame and the model renders the move to a new azimuth, elevation and distance, inventing whatever the camera uncovers; the graph then rebuilds the clip with a music bed through Get Video Components and Create Video, because the model's own score for a near-static shot comes out as a low rumble.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.44.0",
        "capabilities": {
          "workflow": [
            "api",
            "camera"
          ]
        }
      },
      {
        "name": "api_google_gemini_omni_flash_1_2_edit",
        "title": "Gemini Omni 1.2 Flash: Video Edit",
        "task": "Video Edit",
        "model": "Gemini Omni 1.2 Flash",
        "freshness": "new",
        "usage": 0,
        "recommend": "medium",
        "description": "Edit a finished clip instead of regenerating it: keep the shot, timing and audio, and change only what the prompt names. Gemini Omni Flash rewrites the targeted part and re-renders picture and sound together, so the cut still matches.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.44.1",
        "capabilities": {
          "workflow": [
            "api",
            "video-edit"
          ],
          "model_options": {
            "GeminiVideoOmniV2": [
              "Omni Flash 1.1",
              "Omni Flash"
            ]
          }
        }
      },
      {
        "name": "api_google_gemini_omni_flash_1_2_extend",
        "title": "Video Extend: Gemini Omni 1.2 Flash",
        "task": "Video Generation",
        "model": "Gemini Omni 1.2 Flash",
        "freshness": "new",
        "usage": 0,
        "recommend": "medium",
        "description": "Continue a finished clip instead of re-shooting it: the model reads the last frames and keeps the same subject, framing and audio running for the next stretch. A 5-second input becomes a 10-second output without a visible seam.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.44.1",
        "capabilities": {
          "workflow": [
            "api",
            "video-extend"
          ],
          "model_options": {
            "GeminiVideoOmniV2": [
              "Omni Flash 1.1",
              "Omni Flash"
            ]
          }
        }
      },
      {
        "name": "utility_seedvr2_3b_int8_upscale_video",
        "title": "Video Upscale: SeedVR2 3B Int8",
        "task": "Video Upscale",
        "model": "SeedVR2",
        "freshness": "current",
        "usage": 624,
        "recommend": "high",
        "description": "Upscale and restore video footage using SeedVR2 3B Int8, a one-step diffusion model that enhances resolution while maintaining temporal consistency across frames. This workflow takes one degraded video as input and outputs a single high-resolution, restored video. Ideal for restoring old or degraded footage, upscaling low-resolution videos, and enhancing archival media.",
        "io": {
          "inputs": [
            "video: Input video"
          ],
          "outputs": [
            "video: Upscaled video"
          ]
        },
        "minComfyUIVersion": "0.28.0",
        "capabilities": {
          "workflow": [
            "video-upscale",
            "int8"
          ]
        }
      },
      {
        "name": "api_topaz_starlight_precise25",
        "title": "Video Upscale: Topaz Starlight Precise 2.5",
        "task": "Video Upscale",
        "model": "Topaz",
        "freshness": "established",
        "usage": 229,
        "recommend": "medium",
        "description": "Upload a video to enhance its resolution using the Topaz Starlight Precise 2.5 model. This workflow produces sharper 4K output with reduced artifacts compared to previous versions.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.20.2",
        "capabilities": {
          "workflow": [
            "video-upscale",
            "api"
          ]
        }
      },
      {
        "name": "api_wavespeed_flshvsr_video_upscale",
        "title": "Video Upscale: WaveSpeed FlashVSR",
        "task": "Video Upscale",
        "model": "FlashVSR",
        "freshness": "established",
        "usage": 37,
        "recommend": "not_recommended",
        "description": "Upload a video to upscale it to 720p, 1080p, 2K, or 4K using the WaveSpeed FlashVSR model, a fast video super-resolution and restoration model. This workflow enhances detail, reduces artifacts, and ensures consistent motion for a high-quality output. It calls a third-party API, so execution time depends on server-side response.",
        "io": {
          "inputs": [
            "video: Input video to upscale with FlashVSR"
          ],
          "outputs": [
            "video: Output upscaled video with enhanced detail and consistent motion"
          ]
        },
        "minComfyUIVersion": "0.11.0",
        "capabilities": {
          "workflow": [
            "api",
            "video-upscale"
          ]
        }
      },
      {
        "name": "api_topaz_astra2",
        "title": "Video Upscale: Topaz Astra 2 Creative",
        "task": "Video Upscale",
        "model": "Topaz",
        "freshness": "established",
        "usage": 90,
        "recommend": "low",
        "description": "This workflow uses Topaz Astra 2's creative diffusion model to upscale video while adding dynamic detail and texture. You can adjust creativity, sharpness, and use prompt-based guidance for stylized results. It calls a third-party API, so execution time depends on server-side response.",
        "io": {
          "inputs": [
            "video: Input video to upscale with creative enhancement"
          ],
          "outputs": [
            "video: Output upscaled video with enhanced detail and texture"
          ]
        },
        "minComfyUIVersion": "0.20.2",
        "capabilities": {
          "workflow": [
            "api",
            "video-upscale"
          ]
        }
      },
      {
        "name": "utility-gan_upscaler",
        "title": "Video Upscale: Real-ESRGAN",
        "task": "Video Upscale",
        "model": "Real-ESRGAN",
        "freshness": "established",
        "usage": 40,
        "recommend": "not_recommended",
        "description": "Upload a video to upscale it using a fast GAN model. The output resolution is determined by your chosen model (e.g., 2x or 4x). For best results, upscale first, then resize to your target resolution.",
        "io": {
          "inputs": [
            "video: Input video"
          ],
          "outputs": [
            "video: Upscaled video"
          ]
        },
        "minComfyUIVersion": "0.3.31",
        "capabilities": {
          "workflow": [
            "video-upscale",
            "api"
          ]
        }
      },
      {
        "name": "api_bytedance_vcube_video_enhance",
        "title": "Video Upscale: ByteDance vCube",
        "task": "Video Upscale",
        "model": "ByteDance",
        "freshness": "recent",
        "usage": 73,
        "recommend": "low",
        "description": "Enhance video quality with ByteDance vCube, an AI-driven tool that upscales and restores footage through shot-adaptive processing, fixing compression artifacts and color distortion while preserving original texture. Ideal for improving UGC clips, refining AIGC content, and restoring classic films with vibrant, clear results.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.33.4",
        "capabilities": {
          "workflow": [
            "api",
            "video-upscale"
          ]
        }
      },
      {
        "name": "api_topaz_video_enhance",
        "title": "Video Upscale: Topaz Astra Fast",
        "task": "Video Upscale",
        "model": "Topaz",
        "freshness": "established",
        "usage": 59,
        "recommend": "low",
        "description": "Enhance videos with Topaz AI. Supports resolution upscaling using Starlight (Astra) Fast model and frame interpolation with apo-8 model.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.20.2",
        "capabilities": {
          "workflow": [
            "api",
            "video-upscale"
          ]
        }
      },
      {
        "name": "utility_hitpaw_video_enhance",
        "title": "Video Upscale: HitPaw",
        "task": "Video Upscale",
        "model": "HitPaw",
        "freshness": "established",
        "usage": 31,
        "recommend": "not_recommended",
        "description": "Upload a video to enhance and upscale using the HitPaw Video Enhance API. The workflow generates a restored output with improved clarity, temporal stability, and optional portrait enhancement, supporting various resolution targets. This workflow calls a third-party API, so execution time depends on server-side response.",
        "io": {
          "inputs": [
            "video: Input video to enhance and upscale with HitPaw"
          ],
          "outputs": [
            "video: Output enhanced video with improved clarity and stability"
          ]
        },
        "minComfyUIVersion": "0.12.0",
        "capabilities": {
          "workflow": [
            "video-upscale",
            "api"
          ],
          "model_options": {
            "HitPawVideoEnhance": [
              "Portrait Restore Model (1x)",
              "Portrait Restore Model (2x)",
              "General Restore Model (1x)",
              "General Restore Model (2x)",
              "General Restore Model (4x)",
              "Ultra HD Model (2x)",
              "Generative Model (1x)"
            ]
          }
        }
      },
      {
        "name": "api_clarity_ai_crystal_video_upscale",
        "title": "Clarity AI: Crystal Video Upscaler",
        "task": "Video Upscale",
        "model": "Crystal Upscaler",
        "freshness": "new",
        "usage": 74,
        "recommend": "medium",
        "description": "Upscale a video with Clarity AI's Crystal model, running through the Fal proxy: one video in, one upscaled video out. Set scale_factor to multiply the frame size, and the output keeps the source resolution, frame count and audio track. The result is capped at 9 megapixels per frame and 360 megapixel-seconds per clip, so long or oversized clips come back upscaled by less than the factor you asked for. Ideal for restoring generated clips, preparing footage for large screens, and sharpening soft AI video.",
        "io": {
          "inputs": [
            "video: Input video"
          ],
          "outputs": [
            "video: Upscaled video"
          ]
        },
        "minComfyUIVersion": "0.40.0",
        "capabilities": {
          "workflow": [
            "api",
            "video-upscale"
          ],
          "model_options": {
            "ClarityCrystalVideoUpscaleNode": [
              "crystal-video-upscaler"
            ]
          }
        }
      },
      {
        "name": "utility_depth_anything3_video_depth_estimation",
        "title": "Depth Map: Depth Anything 3",
        "task": "Depth Estimation",
        "model": "Depth Anything v3",
        "freshness": "current",
        "usage": 207,
        "recommend": "medium",
        "description": "Upload 1 image to generate a depth map for video depth estimation. Output a single depth-processed image ready for use in downstream workflows. Ideal for video generation tasks requiring depth control, 3D scene preprocessing, and animation depth mapping.",
        "io": {
          "inputs": [
            "image: Input image"
          ],
          "outputs": [
            "image: Depth map"
          ]
        },
        "minComfyUIVersion": "0.25.0",
        "capabilities": {
          "workflow": [
            "controlnet"
          ]
        }
      },
      {
        "name": "utility_video_segment_sam3",
        "title": "Video Segmentation: SAM3",
        "task": "Image Segmentation",
        "model": "SAM3",
        "freshness": "established",
        "usage": 141,
        "recommend": "low",
        "description": "Use the SAM3 model to segment the main subject or content from a video, isolating specific objects or regions. Input a video and receive a segmented mask output.",
        "io": {
          "inputs": [
            "image: Input image"
          ],
          "outputs": [
            "image: Segmentation mask"
          ]
        },
        "minComfyUIVersion": "0.20.0",
        "capabilities": {
          "workflow": [
            "element-segmentation"
          ]
        }
      },
      {
        "name": "api_grok_video_extend",
        "title": "Video Extend: Grok",
        "task": "Video Generation",
        "model": "Grok",
        "freshness": "established",
        "usage": 24,
        "recommend": "not_recommended",
        "description": "Extend a short video by generating a seamless continuation using the fast Grok API from xAI. Upload a source video and a text prompt describing the next scene to produce an extended video output. This workflow calls a third-party API — execution time depends on server-side response.",
        "io": {
          "inputs": [
            "video: Source video to extend",
            "text: Text prompt describing the next scene"
          ],
          "outputs": [
            "video: Extended video output"
          ]
        },
        "minComfyUIVersion": "0.18.2",
        "capabilities": {
          "workflow": [
            "api",
            "video-extend"
          ],
          "model_options": {
            "GrokVideoExtendNode": [
              "grok-imagine-video"
            ]
          }
        }
      },
      {
        "name": "api_seedance2_5_video_extend",
        "title": "Video Extend: Seedance 2.5",
        "task": "Video Generation",
        "model": "Seedance 2.5",
        "freshness": "recent",
        "usage": 93,
        "recommend": "low",
        "description": "Extend an existing video forward or backward by generating new segments that seamlessly preserve the original characters, scenes, and aspect ratio. The source video must be 1.9–30 seconds long, and you specify the duration of the new clip (4–30 seconds, default 5), with output delivered as MP4 or MOV. Ideal for continuing unfinished footage, creating extended narrative sequences, and building longer edit chains.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.33.2",
        "capabilities": {
          "workflow": [
            "api",
            "video-extend"
          ],
          "model_options": {
            "ByteDance2ReferenceNodeV2": [
              "Seedance 2.5",
              "Seedance 2.5 Draft",
              "Seedance 2.0",
              "Seedance 2.0 Fast",
              "Seedance 2.0 Mini"
            ]
          }
        }
      },
      {
        "name": "api_minimax_h3_max_reframe",
        "title": "Aspect Ratio Reframe: MiniMax H3 Max",
        "task": "Video Generation",
        "model": "MiniMax H3",
        "freshness": "new",
        "usage": 0,
        "recommend": "medium",
        "description": "Reframe a clip to a new aspect ratio with MiniMax H3 Max: the source keeps its position, timing and audio, and the model generates the surrounding picture instead of cropping the subject. Choose 16:9, 9:16, 1:1, 4:3, 3:4, 21:9 or 9:21 at 480P to 1080P and guide the new area with a prompt. Ideal for turning a vertical social clip into a horizontal edit, or filling a wide frame from a square source.",
        "io": {
          "inputs": [
            "video: Input video"
          ],
          "outputs": [
            "video: Interpolated video"
          ]
        },
        "minComfyUIVersion": "0.44.0",
        "capabilities": {
          "workflow": [
            "api",
            "outpainting"
          ]
        }
      },
      {
        "name": "api_minimax_h3_max_recast",
        "title": "Cast Swap: MiniMax H3 Max",
        "task": "Video Edit",
        "model": "MiniMax H3",
        "freshness": "new",
        "usage": 18,
        "recommend": "medium",
        "description": "Swap the people in a finished clip with MiniMax H3 Max Recast: connect up to four reference photos and the model rewrites the cast while keeping the motion, the framing, the cuts and the original audio. Clips run 5 to 30 seconds with no shot longer than 15 seconds, at 480P to 1080P, with an option for more natural skin and blending. Ideal for recasting a shot without a reshoot, localising a presenter, or testing a look with a different face.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.44.0",
        "capabilities": {
          "workflow": [
            "api",
            "video-edit"
          ]
        }
      },
      {
        "name": "api_vidu_video_extension",
        "title": "Video Extend: Vidu Q2",
        "task": "Video Generation",
        "model": "Vidu Q2",
        "freshness": "established",
        "usage": 13,
        "recommend": "not_recommended",
        "description": "Use the Vidu Q2 high-quality model to extend a video up to 7 seconds and enhance it to 1080p. The workflow appends generated content seamlessly to the original clip. This utility workflow calls a third-party API; execution time depends on server-side response.",
        "io": {
          "inputs": [
            "video: Input video to extend with Vidu Q2"
          ],
          "outputs": [
            "video: Output extended and upscaled video (up to 7 seconds, 1080p)"
          ]
        },
        "minComfyUIVersion": "0.12.0",
        "capabilities": {
          "workflow": [
            "video-extend",
            "api"
          ],
          "model_options": {
            "ViduExtendVideoNode": [
              "viduq2-pro",
              "viduq2-turbo"
            ]
          }
        }
      },
      {
        "name": "api_google_gemini_omni_flash_1_1_extend",
        "title": "Video Extend: Gemini Omni 1.1 Flash",
        "task": "Video Generation",
        "model": "Gemini Omni 1.1 Flash",
        "freshness": "recent",
        "usage": 39,
        "recommend": "not_recommended",
        "description": "Extend an existing video by up to 10 seconds per step, building stories up to 40 seconds total with Gemini Omni 1.1 Flash. The model analyzes the last 10 seconds of context, keeping characters, motion, and audio coherent as the scene continues from where it left off. Optionally attach reference images to introduce new characters mid-story. Works on uploaded clips and previously generated videos; describe how the scene should continue in plain language. Ideal for longer narratives, scene branching, and iterative storytelling.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.34.2",
        "capabilities": {
          "workflow": [
            "api",
            "video-extend"
          ],
          "model_options": {
            "GeminiVideoOmniV2": [
              "Omni Flash 1.1",
              "Omni Flash"
            ]
          }
        }
      },
      {
        "name": "utility_video_frame_interpolation",
        "title": "Frame Interpolation: FILM or RIFE",
        "task": "Frame Interpolation",
        "model": "FILM",
        "freshness": "established",
        "usage": 59,
        "recommend": "low",
        "description": "Upload a video and select either RIFE or FILM frame interpolation model to generate smoother, higher frame rate video output.",
        "io": {
          "inputs": [
            "video: Input video"
          ],
          "outputs": [
            "video: Interpolated video"
          ]
        },
        "minComfyUIVersion": "0.20.0",
        "capabilities": {
          "workflow": [
            "frame-interpolation"
          ]
        }
      },
      {
        "name": "utility_void_video_inpainting",
        "title": "Video Inpainting: VOID",
        "task": "Video Edit",
        "model": "VOID",
        "freshness": "current",
        "usage": 108,
        "recommend": "low",
        "description": "Upload a video and mask the object you want to remove. Generate a clean video with the object and its physical interactions deleted.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.21.0",
        "capabilities": {
          "workflow": [
            "video-edit",
            "inpainting"
          ]
        }
      },
      {
        "name": "api_bria_video_replace_background",
        "title": "Replace Background: Bria",
        "task": "Video Edit",
        "model": "BRIA",
        "freshness": "current",
        "usage": 11,
        "recommend": "not_recommended",
        "description": "This workflow uses the BRIA API to replace a video’s background with a new image. It accepts a video and a background image as inputs, automatically detects the main subject, and outputs the edited video with the new background. BRIA’s accurate background replacement and transparency handling make it ideal for automated video cleanup and virtual production.",
        "io": {
          "inputs": [
            "image"
          ],
          "outputs": [
            "image"
          ]
        },
        "minComfyUIVersion": "0.24.1",
        "capabilities": {
          "workflow": [
            "api",
            "video-edit",
            "background-removal"
          ]
        }
      },
      {
        "name": "api_bria_remove_video_background_transparent",
        "title": "Remove Background: Bria Transparent",
        "task": "Remove Background",
        "model": "BRIA",
        "freshness": "current",
        "usage": 67,
        "recommend": "low",
        "description": "Upload a video and remove its background via the BRIA API, generating transparent image frames and mask sequences. The workflow outputs a WebM video with an alpha channel for seamless transparency. This is a third-party API call; processing time depends on the server response.",
        "io": {
          "inputs": [
            "video: Input video to extract with transparent background"
          ],
          "outputs": [
            "video: Output WebM video with alpha channel transparency",
            "image: Output image frames with transparent backgrounds",
            "image: Output mask sequences"
          ]
        },
        "minComfyUIVersion": "0.24.1",
        "capabilities": {
          "workflow": [
            "api",
            "background-removal"
          ]
        }
      },
      {
        "name": "api_bria_remove_video_background",
        "title": "Remove Background: Bria Solid Color",
        "task": "Remove Background",
        "model": "BRIA",
        "freshness": "current",
        "usage": 18,
        "recommend": "not_recommended",
        "description": "This workflow uses the BRIA API to remove the background from an input video, outputting a clip with the subject isolated on a solid color or transparent background. BRIA’s video background removal and transparency capabilities drive the extraction. Because it relies on third-party API execution, processing time depends on server-side response.",
        "io": {
          "inputs": [
            "video: Input video to remove background from"
          ],
          "outputs": [
            "video: Output video with background removed"
          ]
        },
        "minComfyUIVersion": "0.14.0",
        "capabilities": {
          "workflow": [
            "api",
            "background-removal"
          ]
        }
      },
      {
        "name": "api_veed_remove_video_background",
        "title": "VEED Remove Video Background: Matte Export",
        "task": "Remove Background",
        "model": "VEED",
        "freshness": "new",
        "usage": 43,
        "recommend": "medium",
        "description": "Cut the subject out of a video with VEED Remove Video Background: the node segments the foreground in every frame and returns the frames plus an alpha mask, and Join Image with Alpha turns that mask into a transparent PNG sequence ready to composite. Choose standard, fast or green-screen keying; clips up to 60 seconds.",
        "io": {
          "inputs": [
            "image: Input image"
          ],
          "outputs": [
            "image: Background removed"
          ]
        },
        "minComfyUIVersion": "0.43.0",
        "capabilities": {
          "workflow": [
            "api",
            "background-removal"
          ]
        }
      },
      {
        "name": "api_veed_dubbing",
        "title": "VEED Dubbing: Video Translation",
        "task": "Video Generation",
        "model": "VEED",
        "freshness": "new",
        "usage": 41,
        "recommend": "medium",
        "description": "Translate the speech in a video into another language with VEED Dubbing: each speaker's voice is cloned and their lips are re-synced to the translation. Pick between the cheaper standard tier and pro, which adds regional accents and stereo audio. Built for footage with a visible face and a talk track.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.43.1",
        "capabilities": {
          "workflow": [
            "api"
          ]
        }
      },
      {
        "name": "api_veed_clean_audio_video",
        "title": "VEED Clean Audio: Video Speech Cleanup",
        "task": "Video Edit",
        "model": "VEED",
        "freshness": "new",
        "usage": 0,
        "recommend": "medium",
        "description": "Clean the speech inside a video with VEED Clean Audio: background noise, music and effects are stripped from the soundtrack, pauses are cleaned completely, and the clip comes back with 48 kHz mono speech, optionally normalized to a target loudness while the picture stays as shot. Video up to 1799 seconds.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.43.0",
        "capabilities": {
          "workflow": [
            "api",
            "video-edit"
          ]
        }
      },
      {
        "name": "api_bria_video_green_screen",
        "title": "Green Screen: Bria",
        "task": "Remove Background",
        "model": "BRIA",
        "freshness": "current",
        "usage": 47,
        "recommend": "not_recommended",
        "description": "Upload a video to generate a green screen version using the BRIA API. The workflow leverages BRIA's accurate background removal to isolate the subject and add a solid green background, ready for compositing. This workflow calls a third-party API; execution time depends on server-side response.",
        "io": {
          "inputs": [
            "video: Input video to convert to green screen"
          ],
          "outputs": [
            "video: Output video with green screen background"
          ]
        },
        "minComfyUIVersion": "0.24.1",
        "capabilities": {
          "workflow": [
            "background-removal",
            "api"
          ]
        }
      },
      {
        "name": "utility-bria_remove_video_background",
        "title": "Remove Background: Bria Local",
        "task": "Remove Background",
        "model": "BRIA",
        "freshness": "established",
        "usage": 9,
        "recommend": "not_recommended",
        "description": "Upload a video to the BRIA API to automatically remove its background, isolating the subject with a transparent or solid background. This third-party workflow processes input video and outputs a clean clip with background removed, leveraging BRIA's accurate video background removal and transparency support.",
        "io": {
          "inputs": [
            "video: Input video to remove background from"
          ],
          "outputs": [
            "video: Output video with background removed"
          ]
        },
        "minComfyUIVersion": "0.14.0",
        "capabilities": {
          "workflow": [
            "api",
            "background-removal"
          ]
        }
      },
      {
        "name": "api_beeble_switchx_video_edit",
        "title": "Video Edit: Beeble SwitchX",
        "task": "Video Edit",
        "model": "SwitchX",
        "freshness": "current",
        "usage": 76,
        "recommend": "low",
        "description": "Upload a source video and a reference image, then describe the desired edits with a text prompt using Beeble SwitchX. The workflow outputs an edited video where masked regions follow your prompt and reference, while unmasked areas preserve original pixels and motion, and includes an alpha matte for compositing. This workflow calls a third-party API, so execution time depends on server response.",
        "io": {
          "inputs": [
            "video: Source video to edit",
            "image: Optional reference image for style, lighting, and background guidance",
            "text: Text prompt describing desired edits and environment"
          ],
          "outputs": [
            "video: Edited video with prompt-guided regions and preserved original areas"
          ]
        },
        "minComfyUIVersion": "0.23.0",
        "capabilities": {
          "workflow": [
            "api",
            "video-edit"
          ]
        }
      },
      {
        "name": "utility_sdpose_ood_video_to_pose_map",
        "title": "Pose Map: SDPose-OOD",
        "task": "Video Generation",
        "model": "SDPose-OOD",
        "freshness": "established",
        "usage": 13,
        "recommend": "not_recommended",
        "description": "Upload a video to extract pose keypoints and generate a pose map. The workflow supports multiple person detection and uses an enhanced SDPose model for accurate whole-body feature extraction.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.16.0",
        "capabilities": {
          "workflow": [
            "pose-map"
          ]
        }
      },
      {
        "name": "utility_sdpose_multi_person_video",
        "title": "Pose Detection: SDPose Multi-Person",
        "task": "Video Generation",
        "model": "SDPose-OOD",
        "freshness": "established",
        "usage": 11,
        "recommend": "not_recommended",
        "description": "Upload a video to detect human poses. Supports detection for both single individuals and multiple people within the same scene",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.19.0",
        "capabilities": {
          "workflow": [
            "pose-map"
          ]
        }
      }
    ]
  },
  {
    "category": "Audio",
    "description": "General-purpose workflow templates for native audio generation, including text-to-audio and other core audio workflows supported out of the box.",
    "templates": [
      {
        "name": "api_mirelo_video_to_sfx",
        "title": "Mirelo SFX 2.0: Video to SFX",
        "task": "Audio Generation",
        "model": "Mirelo",
        "freshness": "new",
        "usage": 186,
        "recommend": "medium",
        "description": "Add picture-locked sound effects to a clip with Mirelo sfx-2.0: the model watches the video and times Foley to on-screen action, keeps speech by default, and saves the SFX track, the speech track and the video carrying the new mix.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.43.0",
        "capabilities": {
          "workflow": [
            "api"
          ]
        }
      },
      {
        "name": "api_mirelo_text_to_sfx",
        "title": "Mirelo SFX 2.0: Text to SFX",
        "task": "Audio Generation",
        "model": "Mirelo",
        "freshness": "new",
        "usage": 98,
        "recommend": "medium",
        "description": "Generate sound effects from a text prompt with Mirelo sfx-2.0: name the sound, its material, space and intensity in one line and get a 48 kHz mono audio file, up to 10 minutes.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.43.0",
        "capabilities": {
          "workflow": [
            "text-to-audio",
            "api"
          ]
        }
      },
      {
        "name": "api_veed_clean_audio",
        "title": "VEED Clean Audio: Speech Cleanup",
        "task": "Audio Editing",
        "model": "VEED",
        "freshness": "new",
        "usage": 15,
        "recommend": "medium",
        "description": "Remove background noise from a speech recording with VEED Clean Audio: music and effects go with the noise, pauses are cleaned completely, and the voice comes back as 48 kHz mono audio, optionally normalized to a target loudness. Audio up to 1799 seconds.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.43.0",
        "capabilities": {
          "workflow": [
            "api",
            "audio-edit"
          ]
        }
      },
      {
        "name": "api_elevenlabs_v4_text_to_speech",
        "title": "Eleven v4: Text to Speech",
        "task": "Text to Speech",
        "model": "ElevenLabs",
        "freshness": "recent",
        "usage": 627,
        "recommend": "high",
        "description": "Generate expressive speech from text with Eleven v4, the quality-focused ElevenLabs model. Direct the performance with audio tags such as [whispers] or [excited] written into the script, or upload a voice sample to clone a voice and narrate the script in it. One audio file is output. Ideal for audiobooks, narration, and character voiceovers.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "audio: Generated speech"
          ]
        },
        "minComfyUIVersion": "0.38.3",
        "capabilities": {
          "workflow": [
            "text-to-speech",
            "api"
          ],
          "model_options": {
            "ElevenLabsTextToSpeech": [
              "eleven_v4",
              "eleven_v4_turbo",
              "eleven_multilingual_v2",
              "eleven_v3"
            ]
          }
        }
      },
      {
        "name": "api_elevenlabs_v4_text_to_dialogue",
        "title": "Eleven v4: Text to Dialogue",
        "task": "Text to Speech",
        "model": "ElevenLabs",
        "freshness": "recent",
        "usage": 32,
        "recommend": "not_recommended",
        "description": "Generate a multi-speaker dialogue from text with Eleven v4: give every entry its own line and its own voice, direct the delivery with audio tags such as [quietly] or [laughs], or clone a voice from a sample and let it take a turn. One audio file is output. Ideal for podcasts, game dialogue, and character conversations.\n",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.38.2",
        "capabilities": {
          "workflow": [
            "text-to-speech",
            "api"
          ],
          "model_options": {
            "ElevenLabsTextToDialogue": [
              "eleven_v3",
              "eleven_v4",
              "eleven_v4_turbo"
            ]
          }
        }
      },
      {
        "name": "audio_minimax_music_3",
        "title": "MiniMax Music 3: Text to Music",
        "task": "Text to Music",
        "model": "MiniMax Music 3",
        "freshness": "recent",
        "usage": 645,
        "recommend": "high",
        "description": "Create full-length songs from text prompts with MiniMax Music 3, generating complete 5-minute tracks with stable structure and high-fidelity audio. Input a structured caption describing lyrics, style, mood, and arrangement, and receive a finished song as output. Ideal for music producers prototyping demos, and content creators needing original background tracks.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.33.1",
        "capabilities": {
          "workflow": [
            "text-to-music"
          ]
        }
      },
      {
        "name": "api_bytedance_seed_audio1_0_t2a",
        "title": "Seed Audio 1.0: Text to Audio",
        "task": "Audio Generation",
        "model": "Seed Audio 1.0",
        "freshness": "current",
        "usage": 84,
        "recommend": "low",
        "description": "Seed Audio 1.0 generates audio from text prompts, producing speech, dialogue, background music, and sound effects in a single file. Describe voices, emotions, and scene details to create multi-speaker audio up to 2 minutes. This workflow uses the text-to-audio API capability.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.27.1",
        "capabilities": {
          "workflow": [
            "text-to-audio",
            "api"
          ],
          "model_options": {
            "ByteDanceSeedAudio": [
              "seed-audio-1.0-multilingual",
              "seed-audio-1.0"
            ]
          }
        }
      },
      {
        "name": "api_bytedance_seed_audio1_0_ta2a",
        "title": "Seed Audio 1.0: Text + Audio to Audio",
        "task": "Audio Editing",
        "model": "Seed Audio 1.0",
        "freshness": "current",
        "usage": 68,
        "recommend": "low",
        "description": "This workflow uses Seed Audio 1.0 to transform an uploaded reference audio clip into new audio based on a text prompt. It leverages the model's audio-to-audio and text-conditioned editing capabilities. Executed via API, the output is a generated audio file matching the prompt and reference style.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.27.1",
        "capabilities": {
          "workflow": [
            "api",
            "audio-edit"
          ],
          "model_options": {
            "ByteDanceSeedAudio": [
              "seed-audio-1.0-multilingual",
              "seed-audio-1.0"
            ]
          }
        }
      },
      {
        "name": "api_bytedance_seed_audio1_0_ti2a",
        "title": "Seed Audio 1.0: Text + Image to Audio",
        "task": "Audio Generation",
        "model": "Seed Audio 1.0",
        "freshness": "current",
        "usage": 13,
        "recommend": "not_recommended",
        "description": "This workflow generates audio from a text prompt and an image reference, using the model’s text-to-audio and audio transformation capabilities to derive voice or style from the image. Execution is via the Seed Audio API.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.27.1",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-audio"
          ],
          "model_options": {
            "ByteDanceSeedAudio": [
              "seed-audio-1.0-multilingual",
              "seed-audio-1.0"
            ]
          }
        }
      },
      {
        "name": "audio_stable_audio_3_medium_base",
        "title": "Stable Audio 3.0 Medium Base",
        "task": "Audio Generation",
        "model": "Stable Audio",
        "freshness": "current",
        "usage": 41,
        "recommend": "not_recommended",
        "description": "Input a short text description of a sound, music, or effect. The workflow expands your prompt with Qwen and generates a stereo audio clip from Stable Audio 3.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.19.2",
        "capabilities": {
          "workflow": [
            "text-to-audio"
          ]
        }
      },
      {
        "name": "audio_stable_audio_3_medium",
        "title": "Stable Audio 3.0 Medium",
        "task": "Audio Generation",
        "model": "Stable Audio",
        "freshness": "current",
        "usage": 183,
        "recommend": "low",
        "description": "Input a short text idea, optional duration, seed, and category. Generate stereo audio (music, SFX, or instruments) using Stable Audio 3 with optional AI-driven text expansion.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.19.2",
        "capabilities": {
          "workflow": [
            "text-to-audio"
          ]
        }
      },
      {
        "name": "api_fishaudio_text_to_speech",
        "title": "Fish Audio: Text to Speech",
        "task": "Text to Speech",
        "model": "Fish Audio",
        "freshness": "recent",
        "usage": 53,
        "recommend": "low",
        "description": "Generate natural, expressive speech from text with Fish Audio S2 Pro, a text-to-speech model supporting 80+ languages and fine-grained control over emotion, tone, and pacing using inline tags like [whisper] or [excited]. Input your text and receive one high-quality audio output, with optional style tags woven directly into the script. Ideal for voiceover production, multilingual content creation, and character dialogue in games or animations.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "audio: Generated speech"
          ]
        },
        "capabilities": {
          "workflow": [
            "api",
            "tts"
          ],
          "model_options": {
            "FishAudioTextToSpeech": [
              "drama-3",
              "s2.1-pro",
              "s1"
            ]
          }
        }
      },
      {
        "name": "api_fishaudio_voice_clone_tts",
        "title": "Fish Audio: Voice Clone to Speech",
        "task": "Text to Speech",
        "model": "Fish Audio",
        "freshness": "recent",
        "usage": 281,
        "recommend": "medium",
        "description": "Clone any voice from a short recording, then use it to speak any script with Fish Audio. Record 2-3 clips of 15-20 seconds, connect them to the Instant Voice Clone node, and the cloned voice feeds straight into text to speech with no training wait. Add emotion tags to control how your cloned voice delivers each line. Great for personalized narration, character voices, and content in your own voice.",
        "io": {
          "inputs": [
            "audio: Reference voice recording",
            "text: Prompt"
          ],
          "outputs": [
            "audio: Generated speech"
          ]
        },
        "minComfyUIVersion": "0.34.0",
        "capabilities": {
          "workflow": [
            "api",
            "tts",
            "voice-cloning"
          ],
          "model_options": {
            "FishAudioTextToSpeech": [
              "drama-3",
              "s2.1-pro",
              "s1"
            ]
          }
        }
      },
      {
        "name": "api_fishaudio_speech_to_text",
        "title": "Fish Audio: Speech to Text",
        "task": "Speech to Text",
        "model": "Fish Audio",
        "freshness": "recent",
        "usage": 9,
        "recommend": "not_recommended",
        "description": "Transcribe audio to text with Fish Audio ASR. Language is detected automatically across 80+ languages, with optional word-level timestamps returned as JSON segments. Use it to create subtitles, transcribe recordings, or generate text prompts from audio clips. A sample audio is pre-loaded so you can test it right away.",
        "io": {
          "inputs": [
            "audio: Audio to transcribe"
          ],
          "outputs": [
            "text: Transcription"
          ]
        },
        "minComfyUIVersion": "0.34.0",
        "capabilities": {
          "workflow": [
            "api",
            "speech-to-text"
          ]
        }
      },
      {
        "name": "api_fishaudio_drama3_dialogue",
        "title": "Fish Audio Drama 3: Two-Speaker Dialogue",
        "task": "Audio Generation",
        "model": "Fish Audio",
        "freshness": "new",
        "usage": 35,
        "recommend": "medium",
        "description": "Generate a multi-speaker scene in one pass with Fish Audio Drama 3: connect one Voice Selector per speaker, mark speaker changes with @Voice1 / @Voice2, and direct the delivery with plain-language [directions] and <whisper> word tags.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.39.3",
        "capabilities": {
          "workflow": [
            "api",
            "tts"
          ],
          "model_options": {
            "FishAudioTextToSpeech": [
              "drama-3",
              "s2.1-pro",
              "s1"
            ]
          }
        }
      },
      {
        "name": "api_fishaudio_drama3_directed",
        "title": "Fish Audio Drama 3: Directed Performance",
        "task": "Audio Generation",
        "model": "Fish Audio",
        "freshness": "new",
        "usage": 139,
        "recommend": "medium",
        "description": "Direct a single Fish Audio Drama 3 performance from the text: shift the delivery mid-line with [plain-language directions] and wrap words with <whisper>...</whisper> tags.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.39.3",
        "capabilities": {
          "workflow": [
            "api",
            "tts"
          ],
          "model_options": {
            "FishAudioTextToSpeech": [
              "drama-3",
              "s2.1-pro",
              "s1"
            ]
          }
        }
      },
      {
        "name": "api_heygen_text_to_speech",
        "title": "HeyGen: Text to Speech",
        "task": "Text to Speech",
        "model": "HeyGen",
        "freshness": "current",
        "usage": 22,
        "recommend": "not_recommended",
        "description": "Convert text input into realistic speech and generate a talking avatar image using HeyGen's API. The workflow leverages lip-sync technology to produce a still image of the speaker with synchronized mouth movements. Ideal for creating animated photo characters from simple text prompts.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.28.1",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-speech"
          ]
        }
      },
      {
        "name": "api_heygen_voice_1",
        "title": "HeyGen Voice 1.0: Voice Clone to Speech",
        "task": "Text to Speech",
        "model": "HeyGen",
        "freshness": "new",
        "usage": 134,
        "recommend": "medium",
        "description": "Clone a voice from a short recording with HeyGen Voice 1.0, then speak any script in that voice. Connect one audio clip to Instant Clone and the resulting voice_id feeds straight into text to speech, with no training wait. Use the expressiveness control to stay close to the recording or push more emotion into the delivery. Ideal for narration in your own voice, character dialogue, and multilingual voiceover.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.40.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-speech",
            "voice-cloning"
          ],
          "model_options": {
            "HeyGenVoiceTextToSpeechNode": [
              "heygen-voice-1",
              "heygen-voice-1-turbo"
            ]
          }
        }
      },
      {
        "name": "api_heygen_voice_1_turbo",
        "title": "HeyGen Voice 1.0 Turbo: Voice Clone to Speech",
        "task": "Text to Speech",
        "model": "HeyGen",
        "freshness": "new",
        "usage": 67,
        "recommend": "medium",
        "description": "Clone a voice from a short recording with HeyGen Voice 1.0 Turbo, then speak any script in that voice. It is the same instant clone as Voice 1.0, run through the turbo model, which renders faster for quick drafts and iteration. Connect one audio clip to Instant Clone and the resulting voice_id feeds straight into text to speech, while SaveText keeps the ID on disk for the next session. Ideal for narration in your own voice, character dialogue, and multilingual voiceover.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.40.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-speech",
            "voice-cloning"
          ],
          "model_options": {
            "HeyGenVoiceTextToSpeechNode": [
              "heygen-voice-1",
              "heygen-voice-1-turbo"
            ]
          }
        }
      },
      {
        "name": "api_sonilo_v2m",
        "title": "Sonilo: Video Soundtrack Generation",
        "task": "Audio Generation",
        "model": "Sonilo",
        "freshness": "established",
        "usage": 67,
        "recommend": "low",
        "description": "Generate synchronized soundtracks from video footage using Sonilo. Input a video and produce music that matches its pacing and emotional cues, outputting a timed audio file.",
        "io": {
          "inputs": [
            "video: Input video used to generate a synchronized soundtrack"
          ],
          "outputs": [
            "audio: Generated soundtrack matched to the input video"
          ]
        },
        "minComfyUIVersion": "0.25.0",
        "capabilities": {
          "workflow": [
            "api",
            "music"
          ]
        }
      },
      {
        "name": "api_sonilo_t2m",
        "title": "Sonilo: Text to Music Generation",
        "task": "Text to Music",
        "model": "Sonilo",
        "freshness": "established",
        "usage": 29,
        "recommend": "not_recommended",
        "description": "Generate high-quality, production-ready music from text prompts using Sonilo. Input descriptive text to create original soundtracks with streaming playback and duration control.",
        "io": {
          "inputs": [
            "text: Text prompt describing the desired music style, mood, and content"
          ],
          "outputs": [
            "audio: Generated music audio from the text prompt"
          ]
        },
        "minComfyUIVersion": "0.25.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-music"
          ]
        }
      },
      {
        "name": "api_sonilo_music1_5",
        "title": "Sonilo Music 1.5: Lyrics to Song",
        "task": "Text to Music",
        "model": "Sonilo",
        "freshness": "new",
        "usage": 0,
        "recommend": "medium",
        "description": "Turn your own words into a finished song with Sonilo Music 1.5: describe the style and the scene, then add a Lyrics: block with [Verse] and [Chorus] tags and the model sings it over full instrumentation. Auto length follows the lyrics, or set the duration yourself.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.44.0",
        "capabilities": {
          "workflow": [
            "api",
            "music",
            "text-to-music"
          ]
        }
      },
      {
        "name": "audio_ace_step1_5_xl_turbo",
        "title": "ACE-Step 1.5XL Turbo: Text to Music",
        "task": "Text to Music",
        "model": "ACE-Step",
        "freshness": "established",
        "usage": 244,
        "recommend": "medium",
        "description": "Generate high-quality music from text prompts using the distilled 4B ACE-Step model. Produces commercial-ready audio in just 8 inference steps without CFG.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.12.0",
        "capabilities": {
          "workflow": [
            "music",
            "text-to-music"
          ]
        }
      },
      {
        "name": "audio_ace_step1_5_xl_sft",
        "title": "ACE-Step 1.5XL SFT: Text to Music",
        "task": "Text to Music",
        "model": "ACE-Step",
        "freshness": "established",
        "usage": 72,
        "recommend": "low",
        "description": "Generate high-quality music from text prompts using a 4B parameter SFT model. Control output with CFG for precise prompt adherence.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.12.0",
        "capabilities": {
          "workflow": [
            "text-to-music"
          ]
        }
      },
      {
        "name": "api_elevenlabs_text_to_speech",
        "title": "ElevenLabs: Text to Speech",
        "task": "Text to Speech",
        "model": "ElevenLabs",
        "freshness": "established",
        "usage": 69,
        "recommend": "low",
        "description": "Input text to generate speech with ultra-realistic voices, or upload a voice sample to clone it for synthesis.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "audio: Generated speech"
          ]
        },
        "minComfyUIVersion": "0.15.0",
        "capabilities": {
          "workflow": [
            "text-to-speech",
            "api"
          ],
          "model_options": {
            "ElevenLabsTextToSpeech": [
              "eleven_v4",
              "eleven_v4_turbo",
              "eleven_multilingual_v2",
              "eleven_v3"
            ]
          }
        }
      },
      {
        "name": "api_elevenlabs_text_to_dialogue",
        "title": "ElevenLabs: Text to Dialogue",
        "task": "Text to Speech",
        "model": "ElevenLabs",
        "freshness": "established",
        "usage": 8,
        "recommend": "not_recommended",
        "description": "Input text and select a voice profile to generate a high-quality, emotionally expressive audio dialogue.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.15.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-speech"
          ],
          "model_options": {
            "ElevenLabsTextToDialogue": [
              "eleven_v3",
              "eleven_v4",
              "eleven_v4_turbo"
            ]
          }
        }
      },
      {
        "name": "api_elevenlabs_speech_to_speech",
        "title": "ElevenLabs: Speech to Speech",
        "task": "Audio Generation",
        "model": "ElevenLabs",
        "freshness": "established",
        "usage": 58,
        "recommend": "low",
        "description": "Upload a source audio file to apply a new voice tone or clone a different voice, generating a modified audio output.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.15.0",
        "capabilities": {
          "workflow": [
            "voice-cloning",
            "api"
          ],
          "model_options": {
            "ElevenLabsSpeechToSpeech": [
              "eleven_multilingual_sts_v2",
              "eleven_english_sts_v2"
            ]
          }
        }
      },
      {
        "name": "api_elevenlabs_voice_isolation",
        "title": "ElevenLabs: Voice Isolation",
        "task": "Audio Editing",
        "model": "ElevenLabs",
        "freshness": "established",
        "usage": 13,
        "recommend": "not_recommended",
        "description": "Upload an audio file containing background noise. Use the ElevenLabs API to isolate and output a clean voice track.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.15.0",
        "capabilities": {
          "workflow": [
            "api",
            "audio-edit"
          ]
        }
      },
      {
        "name": "api_elevenLabs_speech_to_text",
        "title": "ElevenLabs: Speech to text",
        "task": "Speech to Text",
        "model": "ElevenLabs",
        "freshness": "established",
        "usage": 8,
        "recommend": "not_recommended",
        "description": "Upload an audio or video file to transcribe its speech into accurate, editable text.",
        "io": {
          "inputs": [
            "audio: Audio to transcribe"
          ],
          "outputs": [
            "text: Transcription"
          ]
        },
        "minComfyUIVersion": "0.15.0",
        "capabilities": {
          "workflow": [
            "api",
            "speech-to-text"
          ],
          "model_options": {
            "ElevenLabsSpeechToText": [
              "scribe_v2"
            ]
          }
        }
      },
      {
        "name": "api_elevenlabs_text_to_sound_effects",
        "title": "ElevenLabs: Text to Sound Effects",
        "task": "Audio Generation",
        "model": "ElevenLabs",
        "freshness": "established",
        "usage": 113,
        "recommend": "low",
        "description": "Input a text prompt to generate custom sound effects and ambient audio.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.15.0",
        "capabilities": {
          "workflow": [
            "text-to-audio",
            "api"
          ],
          "model_options": {
            "ElevenLabsTextToSoundEffects": [
              "eleven_sfx_v2"
            ]
          }
        }
      },
      {
        "name": "audio_ace_step_1_5_split_llm",
        "title": "ACE-Step 1.5 AI Lyric Song Generation",
        "task": "Text Generation",
        "model": "ACE-Step",
        "freshness": "established",
        "usage": 65,
        "recommend": "low",
        "description": "Input a music style and optional lyrics. Generate a full song with AI-created lyrics and audio.",
        "io": {
          "inputs": [
            "text: User prompt"
          ],
          "outputs": [
            "text: Model response"
          ]
        },
        "minComfyUIVersion": "0.35.2",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-audio"
          ],
          "model_options": {
            "GeminiNodeV2": [
              "Gemini 3.8 Flash",
              "Gemini 3.7 Flash",
              "Gemini 3.5 Flash",
              "Gemini 3.1 Pro",
              "Gemini 3.1 Flash-Lite"
            ]
          }
        }
      },
      {
        "name": "audio_ace_step_1_5_split_4b",
        "title": "ACE-Step 1.5 Music Generation (4B LLM )",
        "task": "Audio Generation",
        "model": "ACE-Step",
        "freshness": "established",
        "usage": 79,
        "recommend": "low",
        "description": "Input a text prompt to generate music. This 4B model version offers stronger audio understanding and composition capabilities compared to smaller variants.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.25.0",
        "capabilities": {
          "workflow": [
            "music",
            "text-to-audio"
          ]
        }
      },
      {
        "name": "audio_ace_step_1_5_split",
        "title": "ACE-Step 1.5 Music Generation Workflow",
        "task": "Audio Generation",
        "model": "ACE-Step",
        "freshness": "established",
        "usage": 34,
        "recommend": "not_recommended",
        "description": "Input a text prompt describing the music style and optional lyrics. Generate a full, high-quality audio song in under 10 seconds on consumer hardware.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.25.0",
        "capabilities": {
          "workflow": [
            "text-to-audio",
            "music"
          ]
        }
      },
      {
        "name": "audio_ace_step_1_5_checkpoint",
        "title": "ACE-Step 1.5 Music Generation AIO",
        "task": "Audio Generation",
        "model": "ACE-Step",
        "freshness": "established",
        "usage": 112,
        "recommend": "low",
        "description": "Input style tags and lyrics to generate a full song. The workflow uses the ACE-Step 1.5 model to produce commercial-grade music in under 10 seconds on consumer hardware.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.12.0",
        "capabilities": {
          "workflow": [
            "music",
            "text-to-audio"
          ]
        }
      },
      {
        "name": "audio_stable_audio_example",
        "title": "Stable Audio 1.0: Text to Audio",
        "task": "Audio Generation",
        "model": "Stability",
        "freshness": "established",
        "usage": 32,
        "recommend": "not_recommended",
        "description": "Generate audio from text prompts using Stable Audio.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.35",
        "capabilities": {
          "workflow": [
            "text-to-audio"
          ]
        }
      },
      {
        "name": "audio_ace_step_1_m2m_editing",
        "title": "ACE Step v1 M2M Editing",
        "task": "Audio Editing",
        "model": "ACE-Step",
        "freshness": "established",
        "usage": 35,
        "recommend": "not_recommended",
        "description": "Edit existing songs to change style and lyrics using ACE-Step v1 M2M.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.35",
        "capabilities": {
          "workflow": [
            "audio-edit"
          ]
        }
      },
      {
        "name": "audio_ace_step_1_t2a_song",
        "title": "ACE Step v1 Text to Song",
        "task": "Audio Generation",
        "model": "ACE-Step",
        "freshness": "established",
        "usage": 12,
        "recommend": "not_recommended",
        "description": "Generate songs with vocals from text prompts using ACE-Step v1, supporting multilingual and style customization.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.35",
        "capabilities": {
          "workflow": [
            "text-to-audio"
          ]
        }
      },
      {
        "name": "audio_ace_step_1_t2a_instrumentals",
        "title": "ACE-Step v1 Text to Instrumentals Music",
        "task": "Audio Generation",
        "model": "ACE-Step",
        "freshness": "established",
        "usage": 13,
        "recommend": "not_recommended",
        "description": "Generate instrumental music from text prompts using ACE-Step v1.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.25.0",
        "capabilities": {
          "workflow": [
            "text-to-audio"
          ]
        }
      }
    ]
  },
  {
    "category": "3D Model",
    "description": "General-purpose workflow templates for native 3D model generation, including image-to-3D and other core 3D workflows supported out of the box.",
    "templates": [
      {
        "name": "3d_pixal3d_trellis2_image_to_model",
        "title": "Pixal3D & TRELLIS.2: Image to Model",
        "task": "Image to 3D",
        "model": "Pixal3D",
        "freshness": "recent",
        "usage": 831,
        "recommend": "high",
        "description": "Upload a single image. Generate a high-fidelity 3D model with PBR textures using either Pixal3D or TRELLIS.2, toggled by a single switch. Both pipelines share the same DINOv3 encoder and shape/texture VAEs; Pixal3D adds MoGe camera estimation for pixel-aligned geometry. Includes background removal and a complete mesh post-processing pipeline: remesh, decimate, UV unwrap, and texture baking.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.34.0",
        "capabilities": {
          "workflow": [
            "image-to-3d",
            "int8"
          ]
        }
      },
      {
        "name": "utility_sam3d_body",
        "title": "SAM 3D Body: 3D Human Mesh Extraction From Video",
        "task": "3D Generation",
        "model": "SAM 3D Body",
        "freshness": "recent",
        "usage": 153,
        "recommend": "low",
        "description": "Extract a full-body 3D human mesh from a single video frame using SAM 3D Body, guided by a text prompt to refine pose and shape estimation. The workflow loads one video and one text prompt, then outputs a reconstructed video showing the recovered human rig. Ideal for motion capture prototyping, biomechanics analysis, and character animation previsualization.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.34.0"
      },
      {
        "name": "api_tripo_p2_image_to_model",
        "title": "Tripo P2: Image to Model",
        "task": "Image to 3D",
        "model": "Tripo P2",
        "freshness": "recent",
        "usage": 152,
        "recommend": "low",
        "description": "Generate a production-ready 3D model from a single reference image using Tripo P2.0 Preview, which builds native quad topology with clean edge flow suited for rigging and animation. It takes 1 input image, with optional multi-view references for better side and back accuracy. Ideal for game characters and props, hard-surface mechanical assets, and high-volume real-time pipelines.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.36.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-3d"
          ]
        }
      },
      {
        "name": "api_tripo_p2_text_to_model",
        "title": "Tripo P2: Text to Model",
        "task": "Text to 3D",
        "model": "Tripo",
        "freshness": "recent",
        "usage": 17,
        "recommend": "not_recommended",
        "description": "Generate production-ready 3D meshes from text prompts with Tripo P2, featuring native quad topology and mesh budgets up to 50,000 triangles or 25,000 quads. Ideal for game characters and props, hard-surface assets, and high-volume pipelines needing clean, rig-ready geometry.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.36.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-3d"
          ]
        }
      },
      {
        "name": "api_tripo_p2_multiview_to_model",
        "title": "Tripo P2: Multiview to Model",
        "task": "3D Generation",
        "model": "Tripo P2",
        "freshness": "recent",
        "usage": 121,
        "recommend": "low",
        "description": "Generate production-ready 3D models from up to four reference views using Tripo P2.0 Preview, which reconstructs geometry with native quad topology and triangle meshes up to 50,000 faces. Ideal for game characters and props ready for rigging, hard-surface mechanical assets, and high-volume real-time pipelines.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.36.0",
        "capabilities": {
          "workflow": [
            "api"
          ]
        }
      },
      {
        "name": "api_meshy7_1_text_to_model",
        "title": "Meshy 7.1: Text to Model",
        "task": "Text to 3D",
        "model": "Meshy",
        "freshness": "recent",
        "usage": 14,
        "recommend": "not_recommended",
        "description": "Generate detailed 3D models from text prompts using Meshy 7.1, with an optional Ultra 4K geometry mode that captures fine surface detail like stitching, wood grain, and scales as real geometry. Ideal for game asset creation, product visualization, and 3D printing preparation.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.36.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-3d"
          ]
        }
      },
      {
        "name": "api_meshy7_1_image_to_model",
        "title": "Meshy 7.1: Image to 3D",
        "task": "Image to 3D",
        "model": "Meshy",
        "freshness": "recent",
        "usage": 125,
        "recommend": "low",
        "description": "Generate a 3D model from a single image using Meshy 7.1's Ultra 4K geometry mode, which produces 4096³ meshes so fine surface detail like stitching, wood grain, and scales becomes real geometry rather than texture. Takes one input image. Ideal for high-fidelity asset creation, detailed prop modeling, and product visualization.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.36.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-3d"
          ]
        }
      },
      {
        "name": "api_meshy_t2_5_smart_topology_t2m",
        "title": "Meshy T2.5: Text to Model (Smart Topology)",
        "task": "Text to 3D",
        "model": "Meshy",
        "freshness": "new",
        "usage": 20,
        "recommend": "medium",
        "description": "Generate a game-ready 3D model from a text prompt with Meshy T2.5. Meshy T2.5 builds the mesh at the triangle budget you set instead of generating a dense mesh and retopologizing it afterwards, so the result drops straight into a game engine. No input files required. Ideal for props, environment dressing and low-poly game assets that need a predictable triangle count.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.42.5",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-3d"
          ]
        }
      },
      {
        "name": "api_meshy_t2_5_smart_topology_i2m",
        "title": "Meshy T2.5: Image to Model (Smart Topology)",
        "task": "Image to 3D",
        "model": "Meshy",
        "freshness": "new",
        "usage": 87,
        "recommend": "medium",
        "description": "Generate a game-ready 3D model from a single image with Meshy T2.5. Meshy T2.5 builds the mesh at the triangle budget you set instead of generating a dense mesh and retopologizing it afterwards, so the result drops straight into a game engine. Texturing with PBR maps is optional and runs at 2k, 4k or 8k. Ideal for turning concept art, product photos and sculpts into engine-ready props.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.42.5",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-3d"
          ]
        }
      },
      {
        "name": "3d_pixal3d_multi_views",
        "title": "Pixal3D: Multi Views to 3D",
        "task": "Image to 3D",
        "model": "Pixal3D",
        "freshness": "recent",
        "usage": 235,
        "recommend": "medium",
        "description": "Upload your own separate view images, or use one mixed multi-view turnaround sheet. The workflow prepares front, back, left and right views, then reconstructs a textured 3D model with Pixal3D Multi-View.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.35.0",
        "capabilities": {
          "workflow": [
            "image-to-3d",
            "int8"
          ]
        }
      },
      {
        "name": "api_hunyuan3d_image_to_model",
        "title": "HY 3D: Image to Model",
        "task": "Image to 3D",
        "model": "Hunyuan3D",
        "freshness": "current",
        "usage": 73,
        "recommend": "low",
        "description": "Upload an image to generate a 3D model with geometry and PBR textures.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.23.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-3d"
          ],
          "model_options": {
            "TencentImageToModelNode": [
              "3.0",
              "3.1"
            ]
          }
        }
      },
      {
        "name": "api_tripo3_1_image_to_model",
        "title": "Tripo H3.1: Image to Model",
        "task": "Image to 3D",
        "model": "Tripo",
        "freshness": "current",
        "usage": 85,
        "recommend": "low",
        "description": "Upload a reference image of your object. Generate a high-detail 3D model with dense geometry and PBR-ready materials.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.34.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-3d"
          ]
        }
      },
      {
        "name": "api_tripo_p1_image_to_model",
        "title": "Tripo P1: Image to Model",
        "task": "Image to 3D",
        "model": "Tripo P1",
        "freshness": "current",
        "usage": 23,
        "recommend": "not_recommended",
        "description": "Upload a reference image of your subject. Generate a game-ready 3D model with clean topology and optimized polygon budgets",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.23.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-3d"
          ]
        }
      },
      {
        "name": "api_tripo_p1_text_to_model",
        "title": "Tripo P1: Text to Model",
        "task": "Text to 3D",
        "model": "Tripo P1",
        "freshness": "current",
        "usage": 7,
        "recommend": "not_recommended",
        "description": "Input a text description of a 3D model. Generate a game-ready 3D model with clean topology and optimized polygon counts",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.23.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-3d"
          ]
        }
      },
      {
        "name": "api_rodin3d_gen2_5_image_to_3d",
        "title": "Rodin Gen2.5: Image to Model",
        "task": "Image to 3D",
        "model": "Rodin",
        "freshness": "current",
        "usage": 57,
        "recommend": "low",
        "description": "This workflow converts a single or multi-view image into a 3D model using Rodin Gen2.5. It accepts a reference image of the object and outputs a GLB model file with adjustable quality and enhanced textures. Execution depends on a third-party API, so overall time varies with server response.",
        "io": {
          "inputs": [
            "image: Reference image of the object to convert to 3D"
          ],
          "outputs": [
            "model: Generated 3D model in GLB format"
          ]
        },
        "minComfyUIVersion": "0.22.2",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-3d"
          ]
        }
      },
      {
        "name": "api_tripo_p1_mv_to_model",
        "title": "Tripo P1: Multiview to Model",
        "task": "Image to 3D",
        "model": "Tripo P1",
        "freshness": "current",
        "usage": 12,
        "recommend": "not_recommended",
        "description": "Upload an image or multi-view references to generate a clean low-poly 3D model with controlled polygon count and organized topology, ready for export to game engines.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.23.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-3d"
          ],
          "model_options": {
            "TripoP1MultiviewToModelNode": [
              "P2"
            ]
          }
        }
      },
      {
        "name": "3d_triposplat_image_to_gaussian_splat",
        "title": "TripoSplat: Image to Gaussian Splat",
        "task": "Image to 3D",
        "model": "TripoSplat",
        "freshness": "current",
        "usage": 198,
        "recommend": "low",
        "description": "Upload a single 2D image. Generate a high-quality 3D Gaussian splat representation with controllable density and budget for rendering.",
        "io": {
          "inputs": [
            "image×4: Multi-view images"
          ],
          "outputs": [
            "model: 3D scene"
          ]
        },
        "minComfyUIVersion": "0.23.0",
        "capabilities": {
          "workflow": [
            "image-to-3d"
          ]
        }
      },
      {
        "name": "3d_idol_image_to_gaussian_splat",
        "title": "Idol: Image to Gaussian Splat",
        "task": "Image to 3D",
        "model": "TripoSplat",
        "freshness": "new",
        "usage": 20,
        "recommend": "medium",
        "description": "Turn a single image into a 3D Gaussian splat with Idol, Comfy-Org's fine-tune of TripoSplat: BiRefNet removes the backdrop, DINOv3 and the Flux2 VAE encode the view, and one pass decodes a 262k-point splat. Exports a .spz splat, an orbit preview and an experimental .glb mesh. Best on stylized characters and props.",
        "io": {
          "inputs": [
            "image×4: Multi-view images"
          ],
          "outputs": [
            "model: 3D scene"
          ]
        },
        "minComfyUIVersion": "0.23.0",
        "capabilities": {
          "workflow": [
            "image-to-3d"
          ]
        }
      },
      {
        "name": "api_meshy7_image_to_model",
        "title": "Meshy 7: Image to 3D",
        "task": "Image to 3D",
        "model": "Meshy",
        "freshness": "recent",
        "usage": 69,
        "recommend": "low",
        "description": "Turn a single image into a production-ready 3D model with Meshy 7, generating a textured mesh that faithfully mirrors your reference's geometry, proportions, and surface details. Ideal for character busts that preserve facial expression, complex mechanical assemblies with correctly placed parts, and dense surface relief like engraved coins or ornate props.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.33.4",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-3d"
          ]
        }
      },
      {
        "name": "api_meshy7_text_to_model",
        "title": "Meshy 7: Text to 3D",
        "task": "3D Generation",
        "model": "Meshy",
        "freshness": "recent",
        "usage": 19,
        "recommend": "not_recommended",
        "description": "Generate production-ready 3D models from a text prompt with Meshy 7, a text-to-3D model that sets a new standard for geometry alignment, faithfully matching the input's proportions, part placement, and surface details. Ideal for concept art turnarounds, character and prop modeling from sketches, and any workflow where preserving the original design's intent is critical.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.33.4",
        "capabilities": {
          "workflow": [
            "api"
          ]
        }
      },
      {
        "name": "api_meshy_image_to_model",
        "title": "Meshy 6: Image to Model",
        "task": "Image to 3D",
        "model": "Meshy",
        "freshness": "established",
        "usage": 17,
        "recommend": "not_recommended",
        "description": "Upload an image and generate a 3D model with Meshy 6. Create characters, objects, or mechanical parts with production-quality geometry and clean topology.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.9.2",
        "capabilities": {
          "workflow": [
            "image-to-3d",
            "api"
          ]
        }
      },
      {
        "name": "api_rodin3d_gen2_5_text_to_3d",
        "title": "Rodin Gen2.5: Text to Model",
        "task": "Text to 3D",
        "model": "Rodin",
        "freshness": "current",
        "usage": 1,
        "recommend": "not_recommended",
        "description": "Use a text prompt to generate a textured 3D model in GLB format via Rodin Gen2.5. This workflow calls a third-party API, so execution speed depends on server-side processing.",
        "io": {
          "inputs": [
            "text: Text prompt describing the desired 3D model"
          ],
          "outputs": [
            "model: Generated 3D model in GLB format"
          ]
        },
        "minComfyUIVersion": "0.22.2",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-3d"
          ]
        }
      },
      {
        "name": "api_meshy_multi_image_to_model",
        "title": "Meshy 6: Multi-Image to Model",
        "task": "Image to 3D",
        "model": "Meshy",
        "freshness": "established",
        "usage": 45,
        "recommend": "not_recommended",
        "description": "Create high-quality 3D models from multiple images using Meshy 6. More photos mean better detail capture, accurate proportions, and cleaner mesh structure.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.23.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-3d"
          ]
        }
      },
      {
        "name": "api_hunyuan3d_part",
        "title": "HY 3D: 3D Parts Decomposition",
        "task": "3D Generation",
        "model": "Hunyuan3D",
        "freshness": "established",
        "usage": 12,
        "recommend": "not_recommended",
        "description": "This workflow decomposes a 3D model (FBX, under 100MB and 30,000 faces) into separated parts using Tencent’s Hunyuan3D model. It outputs a segmented GLB asset ready for reuse and editing. Execution runs via a third-party API, so response time depends on the server.",
        "io": {
          "inputs": [
            "model: Input 3D model in FBX format to decompose into parts"
          ],
          "outputs": [
            "model: Decomposed 3D model in GLB format with separated parts"
          ]
        },
        "minComfyUIVersion": "0.14.0",
        "capabilities": {
          "workflow": [
            "api"
          ]
        }
      },
      {
        "name": "api_hunyuan3d_text_to_model",
        "title": "Hunyuan3D: Text to Model",
        "task": "Text to 3D",
        "model": "Hunyuan3D",
        "freshness": "established",
        "usage": 12,
        "recommend": "not_recommended",
        "description": "Input a text prompt or upload a reference image to generate a detailed 3D model asset.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.23.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-3d"
          ],
          "model_options": {
            "TencentTextToModelNode": [
              "3.0",
              "3.1"
            ]
          }
        }
      },
      {
        "name": "api_hunyuan3d_smart_topology",
        "title": "HY 3D: Smart Topology",
        "task": "3D Generation",
        "model": "Hunyuan3D",
        "freshness": "established",
        "usage": 15,
        "recommend": "not_recommended",
        "description": "Upload a high-poly 3D model file. Generate a lower-polygon, topologically optimized 3D model with a specified reduction level.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.16.4",
        "capabilities": {
          "workflow": [
            "api"
          ]
        }
      },
      {
        "name": "api_tripo3_1_text_to_model",
        "title": "Tripo H3.1: Text to Model",
        "task": "Text to 3D",
        "model": "Tripo",
        "freshness": "current",
        "usage": 11,
        "recommend": "not_recommended",
        "description": "Generate production-ready 3D hero assets with high-density geometry and PBR materials. Takes text prompts or reference images as input and outputs detailed 3D meshes suitable for close-up renders and game assets.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.34.0",
        "capabilities": {
          "workflow": [
            "api",
            "text-to-3d"
          ]
        }
      },
      {
        "name": "api_tripo3_1_multiview_to_model",
        "title": "Tripo H3.1: Multiview to Model",
        "task": "Image to 3D",
        "model": "Tripo",
        "freshness": "current",
        "usage": 60,
        "recommend": "low",
        "description": "Upload multiview images of your object. Generate a high-detail 3D model with dense geometry and PBR-ready materials.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.34.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-3d"
          ]
        }
      },
      {
        "name": "api_tripo3_0_image_to_model",
        "title": "Tripo3.0: Image to Model",
        "task": "Image to 3D",
        "model": "Tripo",
        "freshness": "established",
        "usage": 21,
        "recommend": "not_recommended",
        "description": "Transform images or sketches into 3D models with Tripo 3.0's sharp geometry and production-ready PBR textures.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.35.1",
        "capabilities": {
          "workflow": [
            "image-to-3d",
            "api"
          ]
        }
      },
      {
        "name": "templates_mjm_image_to_3d",
        "title": "Generate 3D Model From Single Image",
        "task": "Image to 3D",
        "model": "Meshy",
        "freshness": "established",
        "usage": 24,
        "recommend": "not_recommended",
        "description": "This workflow converts a single reference image into a textured 3D model by first generating orthographic views via Gemini and Nano Banana, then calling the Meshy API to produce the final model and a video of the scene. Execution depends on third-party API response time.",
        "io": {
          "inputs": [
            "image: Reference image to generate 3D model from"
          ],
          "outputs": [
            "video: Generated video from the 3D scene",
            "model: Generated 3D model with textures"
          ]
        },
        "minComfyUIVersion": "0.35.1",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-3d"
          ],
          "model_options": {
            "GeminiNanoBanana2": [
              "Nano Banana 2 (Gemini 3.1 Flash Image)"
            ]
          }
        }
      },
      {
        "name": "api_meshy_text_to_model",
        "title": "Meshy 6: Text to Model",
        "task": "Text to 3D",
        "model": "Meshy",
        "freshness": "established",
        "usage": 20,
        "recommend": "not_recommended",
        "description": "Generate high-quality 3D models from text prompts using Meshy 6. Create characters with refined geometry, mechanical objects with sharp edges, or game-ready low-poly assets.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.9.2",
        "capabilities": {
          "workflow": [
            "text-to-3d",
            "api"
          ]
        }
      },
      {
        "name": "api_tripo3_0_text_to_model",
        "title": "Tripo3.0: Text to Model",
        "task": "Text to 3D",
        "model": "Tripo",
        "freshness": "established",
        "usage": 8,
        "recommend": "not_recommended",
        "description": "Generate precise 3D models from text with Tripo 3.0's ultra-high resolution geometry and realistic PBR materials.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.35.1",
        "capabilities": {
          "workflow": [
            "text-to-3d",
            "api"
          ]
        }
      },
      {
        "name": "api_tripo_text_to_model",
        "title": "Tripo: Text to Model",
        "task": "Text to 3D",
        "model": "Tripo",
        "freshness": "established",
        "usage": 8,
        "recommend": "not_recommended",
        "description": "Craft 3D objects from descriptions with Tripo's text-driven modeling.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.35.1",
        "capabilities": {
          "workflow": [
            "text-to-3d",
            "api"
          ]
        }
      },
      {
        "name": "api_tripo_image_to_model",
        "title": "Tripo: Image to Model",
        "task": "Image to 3D",
        "model": "Tripo",
        "freshness": "established",
        "usage": 17,
        "recommend": "not_recommended",
        "description": "Generate professional 3D assets from 2D images using Tripo engine.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.35.1",
        "capabilities": {
          "workflow": [
            "image-to-3d",
            "api"
          ]
        }
      },
      {
        "name": "api_tripo_multiview_to_model",
        "title": "Tripo: Multiview to Model",
        "task": "Image to 3D",
        "model": "Tripo",
        "freshness": "established",
        "usage": 22,
        "recommend": "not_recommended",
        "description": "Build 3D models from multiple angles with Tripo's advanced scanner.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.35.1",
        "capabilities": {
          "workflow": [
            "image-to-3d",
            "api"
          ]
        }
      },
      {
        "name": "api_rodin_gen2",
        "title": "Rodin: Gen-2 Image to Model",
        "task": "Image to 3D",
        "model": "Rodin",
        "freshness": "established",
        "usage": 12,
        "recommend": "not_recommended",
        "description": "Generate detailed 4X mesh quality 3D models from photos using Rodin Gen2",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.23.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-to-3d"
          ]
        }
      },
      {
        "name": "api_rodin_image_to_model",
        "title": "Rodin: Image to Model",
        "task": "Image to 3D",
        "model": "Rodin",
        "freshness": "established",
        "usage": 4,
        "recommend": "not_recommended",
        "description": "Generate a 3D model from a single reference photo using Rodin AI, which outputs a model with multiple refinement levels (regular, sketch, detail, smooth). The workflow relies on a third-party API, so execution time depends on server responsiveness.",
        "io": {
          "inputs": [
            "image: Reference photo to convert to 3D model"
          ],
          "outputs": [
            "model: Generated 3D model with multiple refinement levels"
          ]
        },
        "minComfyUIVersion": "0.23.0",
        "capabilities": {
          "workflow": [
            "image-to-3d",
            "api"
          ]
        }
      },
      {
        "name": "api_rodin_multiview_to_model",
        "title": "Rodin: Multiview to Model",
        "task": "Image to 3D",
        "model": "Rodin",
        "freshness": "established",
        "usage": 13,
        "recommend": "not_recommended",
        "description": "Use Rodin's image-to-3D capability to generate a detailed 3D model from multiple photos taken at different angles. This workflow accepts three input images and produces a model with regular, sketch, detail, and smooth refinement levels. It executes via a third‑party API, so response times depend on server availability.",
        "io": {
          "inputs": [
            "image×3: Multi-view photos from different angles for 3D reconstruction"
          ],
          "outputs": [
            "model: Generated 3D model with multiple refinement levels"
          ]
        },
        "minComfyUIVersion": "0.23.0",
        "capabilities": {
          "workflow": [
            "image-to-3d",
            "api"
          ]
        }
      },
      {
        "name": "3d_hunyuan3d-v2.1",
        "title": "HY 3D 2.1: Single View to Model",
        "task": "Image to 3D",
        "model": "Hunyuan3D",
        "freshness": "established",
        "usage": 163,
        "recommend": "low",
        "description": "Reconstruct a 3D model from a single photo using Hunyuan3D 2.1, converting one input image into a raw white-mesh geometry without textures or materials. Ideal for rapid 3D concepting, blockout prototyping, and asset preparation before texturing or retopology.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.28",
        "capabilities": {
          "workflow": [
            "image-to-3d"
          ]
        }
      },
      {
        "name": "3d_hunyuan3d_image_to_model",
        "title": "HY 3D 2.0",
        "task": "Image to 3D",
        "model": "Hunyuan3D",
        "freshness": "established",
        "usage": 51,
        "recommend": "low",
        "description": "Generate a textured-ready 3D mesh from a single image using Hunyuan3D 2.0's shape diffusion pipeline, which aligns geometry to the given image and outputs an untextured white mesh. Ideal for 3D asset prototyping, game and AR content prep, and as a base for later texture painting or handcrafted detailing.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.28",
        "capabilities": {
          "workflow": [
            "image-to-3d"
          ]
        }
      },
      {
        "name": "3d_hunyuan3d_multiview_to_model",
        "title": "HY 3D 2.0 MV",
        "task": "Image to 3D",
        "model": "Hunyuan3D",
        "freshness": "established",
        "usage": 59,
        "recommend": "low",
        "description": "Generate 3D models from multiple views using Hunyuan3D 2.0 MV.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.28",
        "capabilities": {
          "workflow": [
            "image-to-3d"
          ]
        }
      },
      {
        "name": "3d_hunyuan3d_multiview_to_model_turbo",
        "title": "HY 3D 2.0 MV Turbo",
        "task": "Image to 3D",
        "model": "Hunyuan3D",
        "freshness": "established",
        "usage": 29,
        "recommend": "not_recommended",
        "description": "Generate 3D models from multiple views using Hunyuan3D 2.0 MV Turbo.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.28",
        "capabilities": {
          "workflow": [
            "image-to-3d"
          ]
        }
      },
      {
        "name": "api_hunyuan3d_retopo_uv",
        "title": "Hunyuan3D: Model Retopology and UV Unwrapping",
        "task": "3D Generation",
        "model": "Hunyuan3D",
        "freshness": "established",
        "usage": 12,
        "recommend": "not_recommended",
        "description": "Input a dense 3D mesh (from AI generation or photogrammetry) and use Hunyuan3D to produce a lightweight, retopologized GLB model with unwrapped UVs. This workflow optimizes topology and UVs for game asset use. Execution calls a third-party API, so response time depends on the server.",
        "io": {
          "inputs": [
            "model: Input dense 3D mesh (from AI generation or photogrammetry)"
          ],
          "outputs": [
            "model: Retopologized 3D model with UV unwrapped in GLB format"
          ]
        },
        "minComfyUIVersion": "0.16.4",
        "capabilities": {
          "workflow": [
            "api"
          ]
        }
      },
      {
        "name": "api_hunyuan3d_model2uv",
        "title": "HY 3D: UV Unwrapping",
        "task": "3D Generation",
        "model": "Hunyuan3D",
        "freshness": "established",
        "usage": 8,
        "recommend": "not_recommended",
        "description": "Upload a 3D model (FBX, OBJ, or GLB) to perform UV unwrapping via Hunyuan3D. The workflow outputs a processed GLB model with an optimized UV layout for texturing. This task runs through a third-party API, so response time depends on server availability.",
        "io": {
          "inputs": [
            "model: Input 3D model to unwrap UVs for"
          ],
          "outputs": [
            "model: UV-unwrapped 3D model in GLB format"
          ]
        },
        "minComfyUIVersion": "0.14.0",
        "capabilities": {
          "workflow": [
            "api"
          ]
        }
      },
      {
        "name": "3d_moge_panorama_to_mesh",
        "title": "MoGe: Panorama to Mesh",
        "task": "Image to 3D",
        "model": "MoGe",
        "freshness": "current",
        "usage": 31,
        "recommend": "not_recommended",
        "description": "Upload an equirectangular 360° panorama image and generate a textured GLB mesh with vertex colors.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.22.0",
        "capabilities": {
          "workflow": [
            "image-to-3d"
          ]
        }
      },
      {
        "name": "3d_moge_perspective_to_mesh",
        "title": "MoGe: Perspective Geometry Estimation",
        "task": "Image to 3D",
        "model": "MoGe",
        "freshness": "current",
        "usage": 83,
        "recommend": "low",
        "description": "Upload an image to estimate its perspective geometry. Generate a 3D depth map and surface normals from the input.",
        "io": {
          "inputs": [
            "image: Perspective image"
          ],
          "outputs": [
            "model: Estimated mesh"
          ]
        },
        "minComfyUIVersion": "0.22.0",
        "capabilities": {
          "workflow": [
            "image-to-3d"
          ]
        }
      },
      {
        "name": "api_patina_material_t2m",
        "title": "Patina: Text to Material",
        "task": "Text to Music",
        "model": "Patina",
        "freshness": "new",
        "usage": 35,
        "recommend": "medium",
        "description": "Generate a complete, seamlessly tiling PBR material from a text prompt with Patina. One run returns a base texture plus basecolor, normal, roughness, metalness and height maps that line up pixel for pixel, so the whole set drops straight into a 3D material. Works best on structured surfaces such as wood, stone, metal, fabric and ceramic tile; connect an image, optionally with a mask, to repaint an existing surface instead of inventing one.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.43.1",
        "capabilities": {
          "workflow": [
            "api"
          ]
        }
      },
      {
        "name": "api_patina_material_i2m",
        "title": "Patina: Image to Material",
        "task": "3D Generation",
        "model": "Patina",
        "freshness": "new",
        "usage": 17,
        "recommend": "medium",
        "description": "Turn a photo of a surface into a seamless, tiling PBR material with Patina. The source image sets the pattern and Patina redraws it as a tileable material, returning basecolor, normal, roughness, metalness and height maps that line up pixel for pixel. Built for brick, stone, wood, plaster, fabric and ceramic tile; a mask can restrict the redraw to part of the image.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.43.1",
        "capabilities": {
          "workflow": [
            "api"
          ]
        }
      },
      {
        "name": "api_patina_pbr_maps",
        "title": "Patina: PBR Maps",
        "task": "3D Generation",
        "model": "Patina",
        "freshness": "new",
        "usage": 0,
        "recommend": "medium",
        "description": "Turn one image into a set of PBR maps with Patina: basecolor, normal, roughness, metalness and height, all derived from exactly what the picture contains. No regeneration and no tiling - feed a flat, evenly lit surface photo or an existing basecolor map, pick the maps you need and drop them straight into a 3D material.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.43.1",
        "capabilities": {
          "workflow": [
            "api"
          ]
        }
      },
      {
        "name": "api_patina_material_extract",
        "title": "Patina: Material Extract",
        "task": "3D Generation",
        "model": "Patina",
        "freshness": "new",
        "usage": 0,
        "recommend": "medium",
        "description": "Pull a seamless, tiling PBR material out of a photo of a surface with Patina. The prompt says which surface to extract and which structure to keep, and the base texture comes back with basecolor, normal, roughness, metalness and height maps aligned pixel for pixel. Made for photos that already contain the material you want: brick walls, paving, plaster, fabric.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.43.1",
        "capabilities": {
          "workflow": [
            "api"
          ]
        }
      }
    ]
  },
  {
    "category": "Vector",
    "description": "Workflow templates for generating and tracing scalable vector graphics (SVG), including text-to-vector and image-to-vector conversion.",
    "templates": [
      {
        "name": "api_recraft_v4_1_text_to_vector",
        "title": "Recraft 4.1: Text to Vector",
        "task": "Image Generation",
        "model": "Recraft",
        "freshness": "current",
        "usage": 55,
        "recommend": "low",
        "description": "Generate a clean, flat vector logo from a text prompt using Recraft V4.1, outputting a single scalable SVG. No input image required. Recraft delivers sharp, stylistically consistent vector graphics ideal for branding, logo design, and illustrations. This workflow runs via API.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.29.0",
        "capabilities": {
          "workflow": [
            "vector",
            "api"
          ],
          "model_options": {
            "RecraftV4TextToVectorNode": [
              "recraftv4_1_vector",
              "recraftv4_1_utility_vector",
              "recraftv4_1_pro_vector",
              "recraftv4_1_utility_pro_vector",
              "recraftv4",
              "recraftv4_pro",
              "recraftv4_styles_vector",
              "recraftv4_styles_pro_vector"
            ]
          }
        }
      },
      {
        "name": "api_recraft_v4_text_to_vector",
        "title": "Recraft v4: Text to Vector",
        "task": "Image Generation",
        "model": "Recraft",
        "freshness": "established",
        "usage": 18,
        "recommend": "not_recommended",
        "description": "This workflow uses Recraft to convert text prompts into clean, editable vector SVG graphics for logos, icons, or illustrations. It leverages Recraft’s high-quality output and style consistency to produce structured paths suitable for design tools, with API support for Standard and Pro modes.",
        "io": {
          "inputs": [
            "text: Text prompt describing the desired vector graphic"
          ],
          "outputs": [
            "image: Generated SVG vector graphic (logo, icon, or illustration)"
          ]
        },
        "minComfyUIVersion": "0.14.1",
        "capabilities": {
          "workflow": [
            "api",
            "vector"
          ],
          "model_options": {
            "RecraftV4TextToVectorNode": [
              "recraftv4_1_vector",
              "recraftv4_1_utility_vector",
              "recraftv4_1_pro_vector",
              "recraftv4_1_utility_pro_vector",
              "recraftv4",
              "recraftv4_pro",
              "recraftv4_styles_vector",
              "recraftv4_styles_pro_vector"
            ]
          }
        }
      },
      {
        "name": "api_recraft_vector_gen",
        "title": "Recraft: Vector Generation",
        "task": "Text to Image",
        "model": "Recraft",
        "freshness": "established",
        "usage": 39,
        "recommend": "not_recommended",
        "description": "Generate high-quality SVG vector graphics from text prompts using Recraft's AI vector generation model. The workflow accepts a text prompt and outputs an SVG image, with configurable color control. It relies on a third-party API call, so execution time depends on server-side response. Recraft produces sharp, detailed images with Standard and Pro quality modes.",
        "io": {
          "inputs": [
            "text: Text prompt describing the desired vector graphic"
          ],
          "outputs": [
            "image: Generated SVG vector graphic"
          ]
        },
        "minComfyUIVersion": "0.3.34",
        "capabilities": {
          "workflow": [
            "text-to-image",
            "api",
            "vector"
          ]
        }
      },
      {
        "name": "api_quiver_image_to_svg",
        "title": "Arrow 2: Image to SVG",
        "task": "Image Generation",
        "model": "Quiver",
        "freshness": "established",
        "usage": 17,
        "recommend": "not_recommended",
        "description": "Converts an image input into an SVG output.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.39.2",
        "capabilities": {
          "workflow": [
            "api",
            "vector"
          ]
        }
      },
      {
        "name": "api_quiver_text_to_svg",
        "title": "Arrow 2: Text to SVG",
        "task": "Image Generation",
        "model": "Quiver",
        "freshness": "established",
        "usage": 10,
        "recommend": "not_recommended",
        "description": "Input a text prompt and optional reference images to generate one or more SVG vector graphics.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.39.2",
        "capabilities": {
          "workflow": [
            "api",
            "vector"
          ]
        }
      },
      {
        "name": "api_quiver_arrow2_image_to_svg",
        "title": "Arrow2: Image to SVG",
        "task": "Image Generation",
        "model": "Quiver",
        "freshness": "recent",
        "usage": 87,
        "recommend": "low",
        "description": "Convert a raster image into a clean, editable SVG vector using Arrow 2, taking one image and returning one vector output. The model produces cleaner geometry and stronger composition with fewer control points. Ideal for vectorizing sketches, logos, and line art, or preparing scalable brand and product assets.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.39.2",
        "capabilities": {
          "workflow": [
            "api",
            "vector"
          ]
        }
      },
      {
        "name": "api_quiver_arrow2_text_to_svg",
        "title": "Arrow2: Text to SVG",
        "task": "Image Generation",
        "model": "Quiver",
        "freshness": "recent",
        "usage": 13,
        "recommend": "not_recommended",
        "description": "Generate editable SVG vector graphics directly from text prompts using Arrow 2, producing one image output with no file inputs required. Ideal for illustrations, technical diagrams, logo design, and turning written briefs into clean, scalable vector assets.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.39.2",
        "capabilities": {
          "workflow": [
            "api",
            "vector"
          ]
        }
      },
      {
        "name": "api_ideogram_vector_1_text_to_svg",
        "title": "Ideogram Vector 1: Text to SVG",
        "task": "Image Generation",
        "model": "Ideogram",
        "freshness": "new",
        "usage": 53,
        "recommend": "medium",
        "description": "Generate a fully editable SVG from a text prompt with Ideogram Vector 1, optionally guided by up to five reference images, and save it as one vector file. Ideal for logos, posters, flat illustration, patterns and line art that stays crisp at any size and opens directly in a vector editor.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.42.5",
        "capabilities": {
          "workflow": [
            "api",
            "vector"
          ]
        }
      },
      {
        "name": "api_ideogram_vector_1_image_to_svg",
        "title": "Ideogram Vector 1: Image to SVG",
        "task": "Image Generation",
        "model": "Ideogram",
        "freshness": "new",
        "usage": 42,
        "recommend": "medium",
        "description": "Trace a bitmap into a fully editable SVG with Ideogram Vector 1 and save it as one vector file. Ideal for turning logos, flat illustration, patterns and line art into clean paths that stay crisp at any size, with a detail mode that either keeps large flat color blocks or finer color regions. Photographs also work, but convert into many small paths, so clean sources with large areas of flat color and transparent backgrounds vectorize best.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.42.5",
        "capabilities": {
          "workflow": [
            "api",
            "vector"
          ]
        }
      }
    ]
  },
  {
    "category": "Product & Ads",
    "description": "Applied workflows for product shots, placements, UGC-style ads, and commercial sequences. These are purpose-built examples for marketing and merchandising rather than general-purpose generation.",
    "templates": [
      {
        "name": "template_image_speech_to_video",
        "title": "Generate UGC Video With Voice Clone",
        "task": "Lip Sync",
        "model": "ElevenLabs",
        "freshness": "established",
        "usage": 31,
        "recommend": "low",
        "description": "Upload an image to generate AI prompts and synthesize speech. Create a final video with accurate lip sync using ElevenLabs and LTX Video.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.35.2",
        "capabilities": {
          "workflow": [
            "api",
            "lip-sync"
          ]
        }
      },
      {
        "name": "template_product_placement",
        "title": "Product Placement",
        "task": "Image Edit",
        "model": "Gemini-3.1-Pro",
        "freshness": "established",
        "usage": 11,
        "recommend": "low",
        "description": "This workflow seamlessly swaps a product into a reference scene while preserving the original composition and lighting. You provide a reference scene image and a product image; an optional text prompt can guide scene description. The model outputs a product video that naturally integrates the placed item into the scene.",
        "io": {
          "inputs": [
            "image: Reference scene image for product placement",
            "image: Product image to place into the scene",
            "text: Optional text prompt for scene description"
          ],
          "outputs": [
            "video: Product video with the placed item in the scene"
          ]
        },
        "minComfyUIVersion": "0.35.2",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ]
        }
      },
      {
        "name": "template_sferro21_product_ad.app",
        "title": "Product Ad Cinematic Sequence Generator",
        "task": "Image Edit",
        "model": "Nano Banana Pro",
        "freshness": "established",
        "usage": 12,
        "recommend": "low",
        "description": "Upload a product image and a cinematic style reference, then describe your desired ad style with a text prompt. Nano Banana Pro applies its intelligent prompt adherence to generate a stylized product advertisement that preserves the product identity while applying the cinematic look. This workflow calls a third-party API for generation.",
        "io": {
          "inputs": [
            "image: Product image to advertise",
            "image: Cinematic style reference image",
            "text: Text prompt describing the desired ad style"
          ],
          "outputs": [
            "image: Cinematic product advertisement image"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit",
            "style-transfer"
          ],
          "model_options": {
            "GeminiImage2Node": [
              "gemini-3-pro-image-preview",
              "Nano Banana 2 (Gemini 3.1 Flash Image)"
            ]
          }
        }
      },
      {
        "name": "template_eric_exploded_view",
        "title": "Exploded View Animation",
        "task": "FLF2V",
        "model": "Nano Banana 2",
        "freshness": "established",
        "usage": 13,
        "recommend": "low",
        "description": "Upload an image of a device or object to generate a seamless exploded view animation video. The intelligent Google flash model creates start and end frames, then animates the transition to an exploded view and back, outputting a looping video execution.",
        "io": {
          "inputs": [
            "image: Image of a device or object to animate into exploded view"
          ],
          "outputs": [
            "video: Seamless exploded view animation video"
          ]
        },
        "minComfyUIVersion": "0.20.3",
        "capabilities": {
          "workflow": [
            "api",
            "flf2v"
          ]
        }
      },
      {
        "name": "templates-photo_to_product_vid",
        "title": "Phone Image to Product Video: Shoe",
        "task": "Image to Video",
        "model": "Seedream",
        "freshness": "established",
        "usage": 33,
        "recommend": "low",
        "description": "Upload a phone photo of a shoe to generate three studio-grade product videos using the Seedream model via a third-party API. The workflow produces clips showcasing detailed texture, a cinematic pan, and a 360-degree rotation. Seedream's high-quality generation with flexible resolution support delivers professional-looking product visuals.",
        "io": {
          "inputs": [
            "image: Phone photo of a shoe product"
          ],
          "outputs": [
            "video×3: Product videos: detailed texture, cinematic pan, and 360 rotation"
          ]
        },
        "minComfyUIVersion": "0.33.3",
        "capabilities": {
          "workflow": [
            "image-to-video",
            "api"
          ],
          "model_options": {
            "ByteDanceSeedreamNode": [
              "seedream 5.0 pro",
              "seedream 5.0 lite",
              "seedream-4-5-251128",
              "seedream-4-0-250828"
            ]
          }
        }
      },
      {
        "name": "templates-subject_product_swap.app",
        "title": "Swap Product in Character’s Hand, UGC Style",
        "task": "Image Edit",
        "model": "Gemini3 Pro Image Preview",
        "freshness": "established",
        "usage": 19,
        "recommend": "low",
        "description": "Use Google’s Gemini 3 Pro Image Preview model to swap a product held by a character with your brand’s item. Provide two images: the character holding the original product and a standalone shot of the new product. The output seamlessly places your product into the character’s hand, preserving the UGC style. Note: execution uses a third-party API, so response time may vary.",
        "io": {
          "inputs": [
            "image×2: Character holding original product and standalone image of the new product"
          ],
          "outputs": [
            "image: Image with the new product swapped into the character's hand"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ],
          "model_options": {
            "GeminiImage2Node": [
              "gemini-3-pro-image-preview",
              "Nano Banana 2 (Gemini 3.1 Flash Image)"
            ]
          }
        }
      },
      {
        "name": "templates-product_scene_relight",
        "title": "Composite your Product + Scene and Relight",
        "task": "Image Edit",
        "model": "ByteDance",
        "freshness": "established",
        "usage": 13,
        "recommend": "low",
        "description": "Upload a product image and a background scene to blend them realistically using ByteDance's Seedream 4.5 model. The workflow composites the two images and adjusts lighting and shadows for a cohesive product mockup. It executes via a third-party API.",
        "io": {
          "inputs": [
            "image×2: Product image and background scene for compositing and relighting"
          ],
          "outputs": [
            "image: Composited product-in-scene image with matched lighting"
          ]
        },
        "minComfyUIVersion": "0.33.3",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit",
            "relight"
          ],
          "model_options": {
            "ByteDanceSeedreamNode": [
              "seedream 5.0 pro",
              "seedream 5.0 lite",
              "seedream-4-5-251128",
              "seedream-4-0-250828"
            ]
          }
        }
      },
      {
        "name": "templates-car_product",
        "title": "1 Image to Car Product Shots",
        "task": "FLF2V",
        "model": "Kling",
        "freshness": "established",
        "usage": 35,
        "recommend": "low",
        "description": "Upload a photo of your vehicle and generate a studio quality video of the vehicle from multiple angles.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "api",
            "flf2v",
            "image-to-video"
          ],
          "model_options": {
            "ByteDanceSeedreamNodeV3": [
              "seedream 5.0 pro",
              "seedream 5.0 flash",
              "seedream 5.0 lite",
              "seedream-4-5-251128",
              "seedream-4-0-250828"
            ]
          }
        }
      },
      {
        "name": "templates-subject_holding_product.app",
        "title": "Add Product to Character’s Hand, AI UGC",
        "task": "Image Generation",
        "model": "Gemini3 Pro Image Preview",
        "freshness": "established",
        "usage": 10,
        "recommend": "low",
        "description": "Upload a character portrait and a product image; Gemini 3 Pro Image Preview generates a realistic UGC-style image of the character naturally holding the product. This workflow supports brand design, product mockups, and style reference. It uses a third-party API, so response time varies—switch to Nano Banana 2 (Gemini 3.1 Flash Image) for faster results.",
        "io": {
          "inputs": [
            "image×2: Character portrait photo and standalone product image"
          ],
          "outputs": [
            "image: Image of the character holding the product naturally"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "api",
            "portrait"
          ],
          "model_options": {
            "GeminiImage2Node": [
              "gemini-3-pro-image-preview",
              "Nano Banana 2 (Gemini 3.1 Flash Image)"
            ]
          }
        }
      },
      {
        "name": "templates-assemble_dieline",
        "title": "Generate Brand Packaging from Dieline",
        "task": "Image Edit",
        "model": "Google",
        "freshness": "established",
        "usage": 25,
        "recommend": "low",
        "description": "Upload a packaging dieline template and convert it into a realistic 3D package visualization. This workflow uses Google Gemini models (gemini-3-pro-image-preview or Nano Banana 2) via third-party API to intelligently edit the flat template into a fully assembled product mockup. Perfect for brand packaging design previews.",
        "io": {
          "inputs": [
            "image: Packaging dieline template image"
          ],
          "outputs": [
            "image: Assembled 3D package visualization"
          ]
        },
        "minComfyUIVersion": "0.3.71",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ],
          "model_options": {
            "GeminiImage2Node": [
              "gemini-3-pro-image-preview",
              "Nano Banana 2 (Gemini 3.1 Flash Image)"
            ]
          }
        }
      },
      {
        "name": "templates-product_ad-v2.0",
        "title": "Swap Product Into a Reference AD",
        "task": "Style Reference",
        "model": "Gemini3 Pro Image Preview",
        "freshness": "established",
        "usage": 20,
        "recommend": "low",
        "description": "Upload your product image and a reference ad; this workflow generates a static advertisement matching the reference’s visual language. It uses Gemini 3 Pro Image Preview’s style reference capabilities (with optional Gemini 3.1 Flash). Execution depends on third-party API response, ideal for quick brand-aligned mockups.",
        "io": {
          "inputs": [
            "image×2: Product image and reference advertisement for style transfer"
          ],
          "outputs": [
            "image: Product advertisement in the style of the reference ad"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "api",
            "style-reference"
          ],
          "model_options": {
            "GeminiImage2Node": [
              "gemini-3-pro-image-preview",
              "Nano Banana 2 (Gemini 3.1 Flash Image)"
            ]
          }
        }
      }
    ]
  },
  {
    "category": "Character & Fashion",
    "description": "Applied workflows for character sheets, multi-angle portraits, fashion, and identity-consistent looks. These are purpose-built examples for character and apparel work rather than general-purpose generation.",
    "templates": [
      {
        "name": "template_ltx2_3_ic_lora_ingredients",
        "title": "LTX-2.3 IC-LoRA: Reference Sheet Control",
        "task": "Reference to Video",
        "model": "LTX-2.3",
        "freshness": "current",
        "usage": 72,
        "recommend": "low",
        "description": "Generate short video clips that stay visually consistent with a reference sheet inventorying characters, props, and locations. Input a composite reference image and a split prompt describing the action, outputting a single video clip. Ideal for character-driven animation, consistent prop visualization, and location-specific scene generation.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.22.0",
        "capabilities": {
          "workflow": [
            "reference-to-video"
          ]
        }
      },
      {
        "name": "template_seedance2_0_viral_videos_character_swap",
        "title": "Seedance 2.0: Viral Videos Character Swap",
        "task": "Image to Video",
        "model": "GPT Image 2",
        "freshness": "current",
        "usage": 33,
        "recommend": "low",
        "description": "A one click workflow to recreate an input video with your character. All you need to do is upload the video and your character. This workflow uses Google Gemini LLM to analyze the input video and reverse engineer a prompt. You can prompt for specific details to inject into this analysis. The workflow extracts the first frame of the video and swaps your character into it, sends this into Seedance 2.0 Reference-to-Video and generates",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.26.2",
        "capabilities": {
          "workflow": [
            "api",
            "character-replacement",
            "image-to-video"
          ]
        }
      },
      {
        "name": "templates-1_click_multiple_character_angles-v1.0",
        "title": "1 Click Multiple Character Angles",
        "task": "Image Edit",
        "model": "Qwen-Image-Edit",
        "freshness": "established",
        "usage": 249,
        "recommend": "medium",
        "description": "Upload an image of your character and get multiple views of that image from different angles",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.4.0",
        "capabilities": {
          "workflow": [
            "image-edit"
          ]
        }
      },
      {
        "name": "template_contact_sheet-step_1.app",
        "title": "Step 1: Contact Sheet Workflow",
        "task": "Image Edit",
        "model": "Nano Banana Pro",
        "freshness": "established",
        "usage": 41,
        "recommend": "low",
        "description": "Provide character, scene, and product reference images plus a text description. Nano Banana Pro’s superior prompt adherence and image editing composite them into a fashion editorial photo. This workflow executes via a third-party API, so response time depends on server load.",
        "io": {
          "inputs": [
            "image×3: Character reference, scene reference, and product reference images",
            "text: Text description of the desired fashion editorial scene"
          ],
          "outputs": [
            "image: Fashion editorial style composite photograph"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ],
          "model_options": {
            "GeminiImage2Node": [
              "gemini-3-pro-image-preview",
              "Nano Banana 2 (Gemini 3.1 Flash Image)"
            ]
          }
        }
      },
      {
        "name": "template_contact_sheet-step_3.app",
        "title": "Step 3: Contact Sheet Workflow",
        "task": "FLF2V",
        "model": "Gemini-3.1-flash-lite",
        "freshness": "established",
        "usage": 9,
        "recommend": "low",
        "description": "This workflow generates 7 Kling 3.0 FL2V videos from a sequence of 8 keyframe images using Gemini 3.1 Flash Lite. Upload exactly 8 images in order; the model creates smooth transitions between consecutive keyframes. It is Step 3 of a multi-step contact sheet workflow.",
        "io": {
          "inputs": [
            "image×8: Sequence of 8 keyframe images in order"
          ],
          "outputs": [
            "video×7: 7 Kling 3.0 FL2V videos from the keyframes"
          ]
        },
        "minComfyUIVersion": "0.35.2",
        "capabilities": {
          "workflow": [
            "api",
            "flf2v"
          ]
        }
      },
      {
        "name": "template_character_portrait_relighting",
        "title": "Character Portrait Relighting",
        "task": "Relight",
        "model": "Nano Banana Pro",
        "freshness": "established",
        "usage": 33,
        "recommend": "low",
        "description": "Upload a character portrait and a text description of your desired lighting. Nano Banana Pro's superior prompt adherence and intelligent image editing applies cinematic relighting while preserving identity. This workflow calls a third-party API, so execution time depends on server response—ideal for quickly exploring mood and atmosphere.",
        "io": {
          "inputs": [
            "image: Portrait of the character to relight",
            "text: Description of desired lighting conditions"
          ],
          "outputs": [
            "image: Relit portrait with cinematic lighting"
          ]
        },
        "minComfyUIVersion": "0.3.71",
        "capabilities": {
          "workflow": [
            "relight",
            "portrait",
            "api"
          ],
          "model_options": {
            "GeminiImage2Node": [
              "gemini-3-pro-image-preview",
              "Nano Banana 2 (Gemini 3.1 Flash Image)"
            ]
          }
        }
      },
      {
        "name": "templates_rob_portrait_light_migration.app",
        "title": "Portrait Relight",
        "task": "Image Edit",
        "model": "Qwen-Image-Edit",
        "freshness": "established",
        "usage": 42,
        "recommend": "low",
        "description": "Upload your target image and a reference lighting image. Relights your target image.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.4.0",
        "capabilities": {
          "workflow": [
            "image-edit",
            "lora",
            "relight"
          ]
        }
      },
      {
        "name": "templates-character_sheet",
        "title": "360 Full-body Turnaround",
        "task": "Character Reference",
        "model": "Gemini3 Pro Image Preview",
        "freshness": "established",
        "usage": 114,
        "recommend": "low",
        "description": "Upload a character or outfit reference to generate a stitched 360° full-body turnaround sheet with close-up views. Gemini3 Pro Image Preview’s multi-angle support ensures consistent perspectives across the sheet. The workflow runs through a third-party API, so response time depends on server availability.",
        "io": {
          "inputs": [
            "image: Character or outfit reference image"
          ],
          "outputs": [
            "image: Stitched 360 turnaround sheet with full-body and close-up views"
          ]
        },
        "minComfyUIVersion": "0.3.71",
        "capabilities": {
          "workflow": [
            "api",
            "character-reference"
          ],
          "model_options": {
            "GeminiImage2Node": [
              "gemini-3-pro-image-preview",
              "Nano Banana 2 (Gemini 3.1 Flash Image)"
            ]
          }
        }
      },
      {
        "name": "templates-portrait_light_migration",
        "title": "Light Migration for Character Portrait",
        "task": "Relight",
        "model": "Qwen",
        "freshness": "established",
        "usage": 13,
        "recommend": "low",
        "description": "Upload a portrait of your character and a reference lighting image. Generate an image with reference lighting applied and character consistency.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.4.0",
        "capabilities": {
          "workflow": [
            "relight",
            "lora",
            "portrait"
          ]
        }
      },
      {
        "name": "template_eric_seedance_5_subject_and_outfit_combine",
        "title": "Outfit Application for Character Portraits",
        "task": "Virtual Try-On",
        "model": "Seedream 4.5",
        "freshness": "established",
        "usage": 10,
        "recommend": "low",
        "description": "Upload your character portrait and outfit reference via a text prompt describing the outfit and subject. Seedream 4.5 applies its photorealistic image composition and virtual-try-on strengths to generate a 4K portrait with the specific outfit applied. This workflow is executable through the API for consistent wardrobe control.",
        "io": {
          "inputs": [
            "image×3: Character portrait and outfit/clothing reference images"
          ],
          "outputs": [
            "image×2: Generated portraits with the target outfit applied"
          ]
        },
        "minComfyUIVersion": "0.33.3",
        "capabilities": {
          "workflow": [
            "api",
            "virtual-try-on"
          ]
        }
      },
      {
        "name": "templates-1_click_multiple_scene_angles-v1.0",
        "title": "1 click Multiple Scene Angles",
        "task": "Image Edit",
        "model": "Qwen-Image-Edit",
        "freshness": "established",
        "usage": 75,
        "recommend": "low",
        "description": "Upload an image of your scene and generate multiple views of your input scene with 1 click.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.60",
        "capabilities": {
          "workflow": [
            "image-edit"
          ]
        }
      },
      {
        "name": "api_flux_vto",
        "title": "Virtual Try-On: Flux",
        "task": "Virtual Try-On",
        "model": "FLUX",
        "freshness": "current",
        "usage": 19,
        "recommend": "low",
        "description": "Upload a person photo and a garment image for FLUX Virtual Try-On. The fast, customizable model applies the garment onto the person while preserving face and pose, though textures and colors may be less realistic. An optional text prompt can describe the garment. This workflow executes via a third-party API, so response time varies.",
        "io": {
          "inputs": [
            "image×2: Person photo and garment product image for virtual try-on"
          ],
          "outputs": [
            "image: Output image with garment applied to the person"
          ]
        },
        "minComfyUIVersion": "0.23.0",
        "capabilities": {
          "workflow": [
            "api",
            "virtual-try-on"
          ]
        }
      }
    ]
  },
  {
    "category": "Brand & Design",
    "description": "Applied workflows for brand systems, graphic redesign, posters, logos, and layout remixes. These are purpose-built examples for design production rather than general-purpose generation.",
    "templates": [
      {
        "name": "templates_graphic_design_recomposer",
        "title": "Graphic Design Recomposer",
        "task": "Brand Design",
        "model": "Nano Banana 2",
        "freshness": "established",
        "usage": 37,
        "recommend": "low",
        "description": "Upload a graphic design and specify a target aspect ratio. Google's intelligent flash image model recomposes the layout to fill the new dimensions without simple cropping or stretching, preserving design intent. Provide the source image and text description of the desired ratio; the output is a seamlessly adapted graphic in the new format.",
        "io": {
          "inputs": [
            "image: Graphic design image to recompose",
            "text: Desired target aspect ratio"
          ],
          "outputs": [
            "image: Recomposed graphic design in the new aspect ratio"
          ]
        },
        "minComfyUIVersion": "0.20.3",
        "capabilities": {
          "workflow": [
            "api",
            "brand-design"
          ]
        }
      },
      {
        "name": "template_eric_thumbnail_generator",
        "title": "YouTube Thumbnail Generator",
        "task": "Image Edit",
        "model": "Nano Banana Pro",
        "freshness": "established",
        "usage": 9,
        "recommend": "low",
        "description": "This workflow generates a high-contrast YouTube thumbnail from a subject reference image and text prompt. Nano Banana Pro leverages its superior prompt adherence to create dramatic compositions and bold color strategies for maximum engagement. Execution uses a third-party API, so run time depends on server response.",
        "io": {
          "inputs": [
            "image: Subject reference image for the thumbnail",
            "text: Text prompt describing the desired thumbnail concept"
          ],
          "outputs": [
            "image: High-contrast YouTube thumbnail image"
          ]
        },
        "minComfyUIVersion": "0.20.3",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ],
          "model_options": {
            "GeminiNanoBanana2": [
              "Nano Banana 2 (Gemini 3.1 Flash Image)"
            ]
          }
        }
      },
      {
        "name": "templates_doc_workbox_poster_recreator",
        "title": "Realistic Poster Redesign With Character Transformation",
        "task": "Image Edit",
        "model": "Nano Banana Pro",
        "freshness": "established",
        "usage": 10,
        "recommend": "low",
        "description": "Transform uploaded images into realistic poster designs by changing character poses, faces, and clothing while maintaining artistic style and color. Provide an image and text instructions, and Nano Banana Pro’s intelligent prompt adherence handles complex editing. As this workflow calls a third-party API, execution time depends on server response.",
        "io": {
          "inputs": [
            "image: Original poster or image to redesign",
            "text: Content description for the new poster or edit instructions"
          ],
          "outputs": [
            "image: Redesigned poster with character transformation"
          ]
        },
        "minComfyUIVersion": "0.4.0",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit"
          ],
          "model_options": {
            "GeminiImage2Node": [
              "gemini-3-pro-image-preview",
              "Nano Banana 2 (Gemini 3.1 Flash Image)"
            ]
          }
        }
      },
      {
        "name": "templates-3D_logo_texture_animation",
        "title": "Dynamic 3D Logo Animations",
        "task": "Brand Design",
        "model": "ByteDance",
        "freshness": "established",
        "usage": 15,
        "recommend": "low",
        "description": "Upload a vector logo and describe your desired texture; the workflow generates a 3D textured first and last frame, then automatically prompts and produces the full logo animation. ByteDance’s brand design model handles the textured 3D rendering, while its first‑last‑frame‑to‑video capability creates the seamless motion.",
        "io": {
          "inputs": [
            "image: Vector logo image to texture and animate"
          ],
          "outputs": [
            "video: Textured 3D logo animation video"
          ]
        },
        "minComfyUIVersion": "0.35.2",
        "capabilities": {
          "workflow": [
            "api",
            "brand-design",
            "flf2v"
          ]
        }
      },
      {
        "name": "templates-textured_logo_elements",
        "title": "Apply Texture + Elements to Logo",
        "task": "Brand Design",
        "model": "Gemini3 Pro Image Preview",
        "freshness": "established",
        "usage": 6,
        "recommend": "low",
        "description": "Upload a logo, texture reference, and decorative elements. Gemini3 Pro Image Preview composites these into an animated brand asset, then ByteDance Image-to-Video generates the final textured logo video. This workflow preserves brand identity while adding depth, and relies on a third-party API so execution time varies with server response.",
        "io": {
          "inputs": [
            "image×3: Logo image, texture reference, and decorative elements"
          ],
          "outputs": [
            "video: Textured logo animation video"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "api",
            "brand-design",
            "image-to-video"
          ]
        }
      },
      {
        "name": "templates-textured_logotype-v2.1",
        "title": "Apply Texture to Logo",
        "task": "Brand Design",
        "model": "Gemini3 Pro Image Preview",
        "freshness": "established",
        "usage": 13,
        "recommend": "low",
        "description": "Upload your logotype and a texture reference image to generate a short animated brand video. Google's Gemini3 Pro Image Preview transfers the texture for brand design, and the FLF2V model produces the animation. Two images are required: the logo and a texture to apply.",
        "io": {
          "inputs": [
            "image×2: Logotype image and texture reference image"
          ],
          "outputs": [
            "video: Textured logotype animation video"
          ]
        },
        "minComfyUIVersion": "0.8.0",
        "capabilities": {
          "workflow": [
            "api",
            "brand-design",
            "image-to-video",
            "flf2v"
          ]
        }
      },
      {
        "name": "template-multistyle-magazine-cover-nanobananapro",
        "title": "Magazine Cover & Package Design",
        "task": "Image Edit",
        "model": "Nano Banana Pro",
        "freshness": "established",
        "usage": 2,
        "recommend": "low",
        "description": "Design the text layout for your magazine cover photo, and explore packaging options for it.",
        "io": {
          "inputs": [
            "text: Prompt"
          ],
          "outputs": [
            "image: Generated output"
          ]
        },
        "minComfyUIVersion": "0.3.71",
        "capabilities": {
          "workflow": [
            "api",
            "image-edit",
            "brand-design"
          ],
          "model_options": {
            "GeminiImage2Node": [
              "gemini-3-pro-image-preview",
              "Nano Banana 2 (Gemini 3.1 Flash Image)"
            ]
          }
        }
      },
      {
        "name": "templates_3d_match_game_art_style.app",
        "title": "Game Asset Style Transfer Sprite Generator",
        "task": "Style Reference",
        "model": "Nano Banana Pro",
        "freshness": "established",
        "usage": 14,
        "recommend": "low",
        "description": "Generate two clean 2D sprites by supplying an outline and a style reference. The Nano Banana Pro model's intelligent prompt adherence ensures precise art direction matching, while Recraft background removal yields ready-to-use game assets. This workflow calls a third-party API, so execution time depends on server-side response.",
        "io": {
          "inputs": [
            "image: Outline/silhouette reference for the sprite",
            "image: Style reference image for the desired art direction"
          ],
          "outputs": [
            "image×2: Generated 2D sprites matching the specified style and outline"
          ]
        },
        "minComfyUIVersion": "0.9.0",
        "capabilities": {
          "workflow": [
            "api",
            "game",
            "style-reference"
          ],
          "model_options": {
            "GeminiImage2Node": [
              "gemini-3-pro-image-preview",
              "Nano Banana 2 (Gemini 3.1 Flash Image)"
            ]
          }
        }
      }
    ]
  }
]