> ## Documentation Index
> Fetch the complete documentation index at: https://docs.superun.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# AI Video Generation Integration

> Seedance async submit, polling, exception handling, and video persistence workflows.

# AI Video Generation Integration

AI Video Generation Integration builds a full video-generation flow in your product: submit jobs, poll status, handle failures, persist results, and play them in the UI. It fits marketing assets, product demos, content creation, and text-to-video or image-to-video scenarios.

***

## What it can do

* Support text-to-video and image-to-video with reference images.
* Create and track async jobs while avoiding duplicate submits and processing.
* Handle model-service errors and quota limits, with retries.
* Copy results into project file storage for durable playback and sharing links.
* Show queued, generating, success, and failed states in the UI.

<Note>
  Video generation, cloud computing, and file storage are prepared by superun as needed. You usually do not need to manage these dependencies yourself.
</Note>

***

## What to prepare

| Item                         | Notes                                                                      |
| ---------------------------- | -------------------------------------------------------------------------- |
| Generation mode              | Text-to-video, image-to-video, or both                                     |
| Generation parameters        | Model, duration, aspect ratio, and other product requirements              |
| Reference images (if needed) | Must be publicly reachable HTTPS URLs; upload local files to storage first |
| History                      | Whether to keep generation history for review and sharing                  |

***

## Enable in superun

<Steps>
  <Step title="Click Use">
    Open AI Video Generation Integration in the Skill Library and click Use in the top right.
  </Step>

  <Step title="Define parameters">
    Specify text/image mode, model, duration, aspect ratio, and whether to save history.
  </Step>

  <Step title="Verify generation and playback">
    Submit a test job, confirm status display, then play the persisted video from project storage.
  </Step>
</Steps>

***

## Limits

* Generation is async and usually takes about 30 seconds to several minutes. Show clear UI states so users do not resubmit.
* Image-to-video references must be publicly reachable HTTPS URLs; upload local files to storage first.
* Model-returned video URLs are usually temporary and should not be stored as long-term history; copy them into project file storage before playback or sharing.
* Enabling or disabling this skill may change project code and consume compute credits.