> For the complete documentation index, see [llms.txt](https://morphic.com/docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://morphic.com/docs/models/seedance-2.5.md).

# Seedance 2.5

Seedance 2.5 on Morphic generates video from text, images, video clips, and audio. It takes up to 50 references in one generation, runs clips up to 30 seconds at 1080p, and generates its own audio. It can also start from an audio track alone, and edit or extend video you already have.

### How to access Seedance 2.5 on Morphic

1. Open a new or existing file in a project on Morphic
2. Switch the mode to 'Video' in the prompt bar at the bottom, then pick the operation you want: 'Text/Image to Video', 'Frames to Video', 'Reference to Video', or '[Lip Sync](/docs/video/lip-sync.md)'
3. Click the model selector, hover over 'Seedance', and select 'Seedance 2.5'
4. Write your prompt, add your references, and click the blue button in the prompt bar

You can also reach it from [Video focus mode](/docs/video/video-focus-mode.md) by clicking a video on your canvas and selecting it from the Video models dropdown.

### What each operation is for

| Operation           | Use it to                                                         |
| ------------------- | ----------------------------------------------------------------- |
| Text/Image to Video | Generate a clip from a prompt, or animate a single starting image |
| Frames to Video     | Fill the motion between a first and a last frame                  |
| Reference to Video  | Compose from many references, or edit and extend an existing clip |
| Lip Sync            | Drive a character's mouth from a dialogue track                   |

### Task types in Reference to Video

Reference to Video has a 'Task Type' setting that pins what the model does with your inputs. Leave it on Auto unless the model reads your intent wrongly, then set the mode explicitly.

| Task type | What it does                                    |
| --------- | ----------------------------------------------- |
| Auto      | Infers the task from your prompt and references |
| Reference | Generates a new video from the references       |
| Edit      | Edits the input video                           |
| Extend    | Extends the input video                         |

### Editing a video

Edit changes the visuals or audio of a clip you already have, without regenerating it: replace the subject, add or remove an object, redraw part of the frame, or swap the music.

Attach the clip as a reference video and say what to change. Your prompt has to carry the intent in words, so use a verb like edit, add, remove, replace, or change. 'Remove the background music from @Video 1' reads as an edit; 'a car on a coast road' does not.

Two settings are taken out of your hands on an edit, because the output inherits them from the clip you attached:

| Setting      | On an edit                                        |
| ------------ | ------------------------------------------------- |
| Aspect ratio | Inherited from the input video, you cannot set it |
| Duration     | Inherited from the input video, you cannot set it |
| Input length | The clip you attach has to run 4 to 30 seconds    |

The output can land up to 0.4 seconds short of the input, and it can be a non-whole number of seconds.

### Extending a video

Extend continues a clip past its last frame, generates the moment before its first frame, or bridges two clips, keeping the character, rhythm, and style intact.

Attach the clip as a reference video and use a word like extend, continue, or continue the story. Say which direction you mean: 'Extend @Video 1 forward' or 'Continue the 5 seconds before @Video 1'.

Extend keeps the input's aspect ratio, but unlike an edit, you can set how long the new footage runs.

### Specifications

| Specification     | Details                                               |
| ----------------- | ----------------------------------------------------- |
| Supported inputs  | Text, images, video clips, audio files                |
| Max references    | 50 per generation, across images, video, and audio    |
| Max images        | 30                                                    |
| Max video clips   | 10                                                    |
| Max audio clips   | 10                                                    |
| Audio-only input  | Supported, an audio clip alone can drive a generation |
| Output resolution | 480p (SD), 720p (HD), 1080p (FHD)                     |
| Output duration   | 4 to 30s, 5s by default                               |
| Editable clips    | 4 to 30s, output matches the input length             |
| Audio output      | Built-in, and can be switched off                     |

### Prompting tips

* Name each reference in your prompt (@Image 1, @Video 1, @Audio 1) and say what it is for, so the model does not swap their roles.
* Set Task Type to Edit or Extend when you already have a clip. Auto occasionally reads an edit as a fresh generation, and setting it explicitly catches a bad setup before the generation runs rather than after.
* Match the wording to the task type you picked. The model still reads your prompt for intent, so an Edit whose prompt never asks for a change, or an Extend that never says continue, can fail even with the right setting.
* Keep camera and lighting direction to one or two concrete cues; stacked cues cancel each other out.
* For clips past 10 seconds, write the action as beats rather than one long sentence.

### Credits

Seedance 2.5 uses adaptive pricing: the rate depends on your plan, and the generate button shows an estimate before you run it. Attaching a video changes how you are billed, because the footage you attach is charged alongside the clip you generate. See [Plans & Credits](/docs/pricing/plans-and-credits.md) for the rates.

***

If you have any trouble or have any questions, feel free to reach out. We're happy to help. Write to us at [**support@morphic.com**](mailto:support@morphic.com)**.**


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://morphic.com/docs/models/seedance-2.5.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.