Skip to main content
The video generation endpoint supports mixed text, image, video, and audio inputs. Dreamina Seedance 2.0 / Mini / Fast all use the same /v1/videos/generations endpoint.

​
Supported inputs

Audio is not recommended as the only input. Pair it with an image or video. For first-frame and last-frame scenarios, explicitly set role=first_frame and role=last_frame.

​
Asset library

If you want to reuse public assets, first use Create asset, then pass the returned uri into the video task.

​
Prompting tips

  • Describe the subject, motion, camera movement, and mood clearly.
  • When multiple assets are used, refer to them in the prompt as “image 1”, “image 2”, “video 1”, or “audio 1”.
  • For simple text-to-video or image-to-video cases, you can also pass public HTTPS URLs or Base64 directly in content.

​
Recommendations

  1. dreamina-seedance-2-0 for general use.
  2. dreamina-seedance-2-0-mini for lighter tasks.
  3. dreamina-seedance-2-0-fast when speed matters most.
  4. Use /v1/models as the source of truth for exact limits.