# Seedance 2.5 now available on Vercel AI Gateway

> Source: <https://vercel.com/changelog/seedance-2-5-now-available-on-vercel-ai-gateway>
> Published: 2026-08-06 00:00:00+00:00

[Seedance 2.5](https://vercel.com/ai-gateway/models/seedance-2.5) 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 supports text, video, image, and audio as inputs in the same request, following a subject's appearance from one asset and the motion or camera work from another. On a finished video it edits locally, swapping backgrounds, products, or characters while the frame, camera, and pacing stay put, which is how one cut becomes several versions. Prompts work in more than ten languages.

Set `model`

to `bytedance/seedance-2.5`

and call `generateVideo`

from the [AI SDK](https://ai-sdk.dev/):

Pass in image URLs under `referenceImages`

and point at them in the prompt with `[Image 1]`

, `[Image 2]`

.

`referenceVideos`

is numbered separately from the images, so `[Video 1]`

and `[Image 1]`

can appear in the same prompt: motion from one asset, appearance from another.

A reference feeds a new generation, contributing motion, camera work, or pacing to something the model builds from scratch. Editing changes a finished video in place. Mixing the two roles in one prompt gives less stable results, so be explicit about what each source asset is for.

For more details, and to try Seedance 2.5 playground and API reference, go to the [model page](https://vercel.com/ai-gateway/models/seedance-2.5). For more video models on AI Gateway, [view all video models](https://vercel.com/ai-gateway/models?type=video).