BytePlus Seedance API for controllable video generation
Seedance 2.5 renders native 30-second video from a prompt, a still image or a reference clip - with up to 50 reference assets per request and three input modes behind one API.
Credit-based pricing - quota and concurrency tiered by plan - one account for every model.
What the Seedance API gives you
One model family for text-to-video, image-to-video and video-to-video, sized for production traffic rather than one-off demos.
Native 30-second generation
Render 30 seconds of video in a single request instead of stitching shorter clips together.
Up to 50 reference assets
Mix images, video and audio references in one request to steer subject, style and motion.
Three input modes
Drive the model from a text prompt, a still image or an existing clip with the same API surface.
Quota and concurrency tiers
Plan tiers set the throughput, quota and concurrency available to your integration.
One balance for every model
Seedance, Seedream and the rest of the platform share one account and one credit balance.
Same model in API and workbench
Prototype in the browser, then move the same model into your own backend.
Seedance API FAQ
Duration, reference assets, input modes and pricing for the Seedance video API.
Walk through the full workflow in the Seedance 2.5 API guide, read the documentation, or work with stills first in the Seedream API.
Start building with the Seedance API
Pick a token package, generate a key and send your first video request.