Current live workflow
- Seedance 2.5: 4–30 seconds
- Seedance 2.0: 4–15 seconds
- Cdance Video Generator: 5, 10, or 15 seconds
- Native audio and resolution controls shown when supported
Prompt to motion
Turn a written direction into a video
Describe the subject, action, setting, camera, light, and sound. Cdance shows the active model, duration choices, aspect ratio, and estimated credit cost before generation.
Write better directions
A reliable prompt names the subject and action first, then adds the environment, composition, camera movement, lighting, pace, and any sound intent.
- Keep one clear primary action
- Use concrete camera language
- State a vertical or landscape publishing goal
Know before you generate
Output can vary between runs. Review the model and credit estimate in the editor; failed provider tasks are returned as credits automatically by the application workflow.