Seedance 1.0

Seedance 1.0, unveiled by ByteDance's Seed team in June 2025, is a text-to-video and image-to-video model that ranked first worldwide on Artificial Analysis in both text-to-video and image-to-video, ahead of Google's Veo 3 and OpenAI's Sora at the time. Its signature capability is native multi-shot storytelling: up to 10 seconds of 1080p video at 24fps with 2–3 camera cuts, switching between far, mid, and close shots while holding character and scene consistency. Five seconds of Full HD generates in about 41 seconds, notably faster than rivals.

The honest picture: Seedance 1.0 launched without audio generation (fixed in 1.5 Pro's joint audio-video release, Dec 2025, and the 2.x series). In February 2026 the MPA, Disney, and Paramount issued copyright takedown notices, briefly constraining global rollout until IP safeguards and watermarking were added. Access is primarily through ByteDance surfaces (Jimeng/Dreamina, Doubao, CapCut) and the BytePlus ModelArk enterprise API — there is no open-weights release, and the API was still marked "coming soon" well into 2026.