Seedance 2.5

ByteDance's newest video model, announced on 23 June 2026 at the Volcano Engine FORCE conference in Beijing. Seedance 2.5 generates up to 30 seconds of native 4K video from a single prompt, accepts up to 50 multimodal reference inputs, and processes audio in the same latent space as the picture so sound lands on the action rather than being layered over it afterwards. ByteDance skipped 2.1 through 2.4 to mark it as a generational step rather than a point release. It is in enterprise beta with a public launch through ByteDance's own apps, and is not available through MotionArt yet — the moment an API endpoint exists, this page becomes a working generator. Until then, Seedance 2.0 is the newest Seedance you can run here.