Seedance 2.5 now available on Vercel AI Gateway
By Steven Van ·
Seedance 2.5 from ByteDance is now available on AI Gateway.
ByteDance's Seedance 2.5 is now available through Vercel's AI Gateway. It generates clips up to 30 seconds long with synchronised audio, and accepts up to 50 reference inputs in a single request across images, video, audio, and style. Short clips can be extended while carrying over character, scene, and camera movement, and a single clip holds camera movement and continuity without stitching shots together afterward. Prompts work in more than ten languages.
The model can also edit a finished video in place, swapping backgrounds, products, or characters while the frame, camera, and pacing stay fixed, so one shoot can yield several variants. It's called through the AI SDK by setting the model to bytedance/seedance-2.5 and calling generateVideo. Reference assets are passed via an inputReferences array and tagged by media type; a prompt can mix video and image references in the same request, for example pulling camera work from a clip and a subject's appearance from a still, and whether the model builds a new video or edits a supplied one depends on how the prompt is phrased.
Full details are on the model page.
