Seedance 2.0 is ByteDance's latest AI video model, available on Buzzy. It turns text prompts, images, video clips, and audio into cinematic shots of up to 30 seconds at 4K, with sound generated natively alongside the picture.

Four upgrades matter most: a single continuous segment stretched from 5s to 30s, reference inputs raised to 50, region-level editing that repaints part of a clip without re-rendering it, and noticeably tighter prompt adherence.

Any length between 4 and 30 seconds, at 480p, 720p, 1080p, or 4K, in 16:9, 9:16, 1:1, 4:3, 3:4, or 21:9. Longer and higher-resolution generations cost more credits.

Yes. Dialogue, music, and sound effects are produced together with the video rather than added afterwards, including lip-sync across 8+ languages for the full length of the clip.

Up to 50 reference inputs in a single generation, mixing images, video clips, and audio — use them to lock down characters, products, style, camera motion, and sound at the same time.

Yes. Character identity, lighting, and scene layout stay consistent through the whole 30-second take, and reusing the same character reference keeps them stable across separate generations too.

Describe the subject, the camera move, and the lighting in one prompt instead of listing keywords, then attach references for anything that must match exactly. Start at a lower resolution to iterate cheaply, and switch to 4K once the shot is right.

Enter a prompt in the box at the top of this page — no waitlist and no install. New accounts start on the free tier, and paid plans add credits for longer, higher-resolution generations.