# Krea 3 Image

Krea 3 Image edits images with Krea 3, or generates a new image from a text prompt when no image is connected. Connect up to 8 images and describe what to change and what to keep, referring to them in the prompt as @Image1, @Image2, and so on. The aspect ratio and resolution set the output size, and the thinking mode trades speed for a more detailed plan.

## Inputs

### Common Inputs

| Parameter | Description | Data Type | Required | Range |
| --- | --- | --- | --- | --- |
| `model` | The Krea 3 model to use. Selecting a model reveals the parameters it accepts. | DYNAMIC_COMBO | Yes | `"Krea 3"` |

### Krea 3 Inputs

These inputs appear when `model` is set to `Krea 3`.

| Parameter | Description | Data Type | Required | Range |
| --- | --- | --- | --- | --- |
| `model.images` | The images to edit, numbered in connection order. Each connected image counts towards the limit of 8. Leave empty to generate from text only. | IMAGE | No | 0 to 8 images |
| `model.prompt` | What to change and what to keep. Refer to the images as `@Image1`, `@Image2`, and so on, or as image 1, image 2, ... The prompt cannot be empty. | STRING | Yes | - |
| `model.aspect_ratio` | The aspect ratio of the result. `"auto"` lets Krea pick the shape from the connected images and the prompt. (default: `"auto"`) | COMBO | Yes | `"auto"`<br>`"1:1"`<br>`"4:5"`<br>`"3:4"`<br>`"2:3"`<br>`"9:16"`<br>`"5:4"`<br>`"4:3"`<br>`"3:2"`<br>`"16:9"`<br>`"21:9"` |
| `model.resolution` | The output resolution. `"2K"` takes longer and produces a larger image than `"1K"`. (default: `"2K"`) | COMBO | Yes | `"1K"`<br>`"2K"` |
| `model.thinking_mode` | How much work Krea 3 puts into planning the edit. `"fast"` is recommended, while `"high"` plans complex edits in more detail and takes longer. `"ultra"` generates several candidates and returns the best one; it needs at least one image, always outputs 2K and does not support `precise_edit`. (default: `"fast"`) | COMBO | Yes | `"fast"`<br>`"high"`<br>`"ultra"` |
| `model.precise_edit` | Keep everything outside the edited area unchanged. Requires at least one connected image. (default: False) | BOOLEAN | Yes | false<br>true |
| `model.seed` | The seed for the request (default: 42). Krea rewrites the prompt before generating, so the same seed does not repeat an image. | INT | Yes | 0 to 2147483647 |

A reference such as `@Image3` is only accepted when at least three images are connected, otherwise the node raises an error. Reference images are uploaded before the request is sent.

## Outputs

| Output Name | Description | Data Type |
| --- | --- | --- |
| `image` | The generated or edited image. | IMAGE |

> This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! [Edit on GitHub](https://github.com/Comfy-Org/embedded-docs/blob/main/comfyui_embedded_docs/docs/Krea3ImageNode/en.md)

---
**Source fingerprint (SHA-256):** `9e8d036f1ebca81b85e6ef42a015e7e2925bf467fa32e82180ee1b76dcc91c0b`
