AI API Lab

WORKBENCH / VIDEO GENERATION

Seedance Replicate · ByteDance

Generate video from text or first/last frame images with your Replicate token.

02

Describe your video

Describe the subject, motion, camera, lighting, and sound. Put spoken dialogue in double quotes.

Use publicly accessible HTTPS image URLs. A last frame requires a first frame. Seedance 2.5 image mode uses adaptive aspect ratio.

Leaving this page does not cancel a submitted prediction. Keep its ID or check your Replicate predictions before submitting again.

03

Video & prediction

▶

Your generated video will appear here.

Replicate API outputs expire after about one hour. Download completed videos promptly.

Replicate predictions ↗

EXPLORE / HOW TO USE

Seedance video generation with Replicate

Generate video with Seedance 2.0 and 2.5 using your own Replicate API token. Try text or first/last frame images, synchronized audio, and video settings.

What you can try

  • Choose Seedance 2.0 or 2.5 and adjust duration, resolution, and aspect ratio.
  • Generate from text or publicly accessible first/last frame image URLs, with optional synchronized audio.
  • Track and cancel asynchronous predictions, restore them by ID, and download completed videos.
  • Requests use your provider’s API quota and may incur charges. Check the key, model access, and billing if you receive an authentication or quota error.

How to get started

  1. Create a Replicate API token and check your account billing and model access.
  2. Describe the scene, motion, camera, and sound. Optionally enter first/last frame image URLs.
  3. Submit once and check prediction status. If submission fails, check your Replicate predictions before resubmitting.
  4. Load and download the completed video promptly; Replicate API outputs expire after about one hour.
  5. Example: a fox walks through a sunlit forest while the camera tracks slowly. Start with a five-second video and compare camera instructions using the same settings.
Explore all playgrounds →Replicate API keys ↗