Seedance 2.5 now available on Vercel AI Gateway
Sora competitor Seedance 2.5 lands on Vercel's AI Gateway—text, video, image, audio inputs in one request, 30-second clips with native camera continuity.
Why it matters
A capable video generation model (ByteDance's Seedance) is now accessible via a major cloud platform's developer API layer. Practitioners building video features can now integrate it without direct model access; the real news is availability and developer ergonomics, not the model's capabilities themselves.
The key facts
14 to knowSeedance 2.5 available on Vercel AI Gateway
Generates up to 30 seconds in a single clip with native camera movement and continuity
Supports text, video, image, and audio as inputs in same request
Can reference appearance from one asset and motion/camera from another
Local editing: swap backgrounds, products, characters while preserving frame/camera/pacing
Prompts work in 10+ languages
Accessible via Vercel AI SDK and API reference
Generates up to 30 seconds per clip without shot stitching
Supports text, video, image, and audio inputs in single request
Multimodal reference system: can pull appearance from one asset, motion/camera from another
Video editing capability: swap backgrounds, products, characters while preserving frame/camera/pacing
Prompts support 10+ languages
Accessible via SDK, API reference, and playground
ByteDance (parent company) provides the model
Go to the source
Vercel Blogvercel.com
Publisher excerpt: from ByteDance is now available on AI Gateway. It generates up to 30 seconds in a single clip, holding camera movement and continuity without stitching shots together in post. Short clips can also be extended with character, scene, and camera movement carried over.Seedance 2.5 Seedance 2.5 supports…