# MiniMax H3 Max Extend Video

This node continues a video with new footage guided by a prompt and optional audio references. The source video is uploaded, the continuation is generated, and the result is returned either as the whole re-rendered video or as the new footage alone. Two models are available: "MiniMax H3 Max" and "MiniMax H3 Max Turbo", which is cheaper per unit.

## Inputs

### Common Inputs

| Parameter | Description | Data Type | Required | Range |
| --- | --- | --- | --- | --- |
| `model` | Model to use for continuing the video. Both models expose the same inputs below. | DYNAMIC_COMBO | Yes | "MiniMax H3 Max"<br>"MiniMax H3 Max Turbo" |

### Model Inputs

These inputs apply to both models.

| Parameter | Description | Data Type | Required | Range |
| --- | --- | --- | --- | --- |
| `video` | Video to extend, 1.625 to 60 seconds long, up to 50 MB. The aspect ratio must be between 0.4 and 2.5. | VIDEO | Yes | 1.625 to 60 seconds |
| `prompt` | What happens next. Describe the continuation, not the source. Refer to the reference audios as `@Audio1`..`@Audio3`, for example "@Audio1 for the voice". | STRING | Yes | 1 to 50000 characters |
| `reference_audios` | Growable slot: connect up to 3 audio clips (`audio_1`...`audio_3`) for the new footage, at least 2 seconds each and 14.5 seconds in total. | AUDIO | No | 0 to 3 clips |
| `duration` | Seconds of new footage to add (default: 5). | INT | Yes | 1 to 15 |
| `resolution` | Resolution of the output video (default: "768P"). | COMBO | Yes | "480P"<br>"768P"<br>"1080P"<br>"2K" |
| `aspect_ratio` | Aspect ratio of the output. "auto" follows the source video; the other values crop the video to the selected ratio (default: "auto"). | COMBO | Yes | "auto"<br>"21:9"<br>"16:9"<br>"4:3"<br>"1:1"<br>"3:4"<br>"9:16" |
| `output` | "extended" returns the whole video, re-rendered at the selected resolution, with the new footage. "continuation" returns only the new footage, starting with a 1.6 second repeat of the source ending (default: "extended"). | COMBO | Yes | "extended"<br>"continuation" |
| `prompt_expansion` | Rewrite the prompt using the source video to keep the continuation consistent. Turn it off to get reproducible results for the same seed (default: true). | BOOLEAN | No | true<br>false |
| `seed` | Random seed. The same request with the same seed gives similar, but not guaranteed identical, results (default: 42). | INT | Yes | 0 to 4294967295 |

### Parameter Constraints

- At 2K resolution the "extended" output is limited to 45 seconds in total, source video plus new footage. Use "continuation" or a lower resolution for longer clips.
- Each reference audio clip must be at least 2 seconds long, and all of them together cannot exceed 14.5 seconds.

## Outputs

| Output Name | Description | Data Type |
| --- | --- | --- |
| `video` | The extended video or the new footage alone, depending on the `output` setting. | VIDEO |

> 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/MinimaxHailuo03MaxExtendVideoNode/en.md)

---
**Source fingerprint (SHA-256):** `db41a14db392fa450689c8d0d73d38b4ce9ea59ee30cfe4d6d2b0501cf0fc9a8`
