1. Introduction
Seedance 2.0 is a ModelVia model.Seedance 2.0 is a ByteDance Seed video model for high-quality text-to-video, image-to-video, and reference-guided clips with detailed motion and camera control.
This page combines the model overview, interactive testing surface, request schemas, and production API workflow in one place.
2. Key Features & Innovations
- • Cinematic clips
3. Model Architecture & Technical Details
Seedance 2.0 generates high-quality short videos from text, images, or reference media with detailed motion and camera control.
Describe the subject, motion, camera, and style.
Choose the Seedance reference workflow.
Optional starting frame for first/last frame generation.
Optional ending frame for first/last frame generation.
Optional reference images for reference-guided generation.
Optional reference videos for omnidirectional reference generation.
4. Performance Highlights
- • Poll the prediction endpoint until complete.
