Seedance AI

Seedance 2.0 is ByteDance Seed's native multimodal audio-video generation model. It accepts text, image, video, and audio inputs and can use them as references for video generation, extension, and editing.

Yes. Seedance 2.0 generates audio and video together. Depending on the prompt and references, the result can include dialogue, sound effects, ambience, music, and dual-channel audio.

Seedance 2.0 supports outputs from 4 to 15 seconds. It can generate a connected multi-shot sequence within that duration.

You can combine natural-language instructions with images, video clips, and audio clips. This interface accepts up to 12 reference files in total, with per-type limits of 9 images, 3 videos, and 3 audio files. References can guide the subject, composition, action, camera movement, visual style, effects, timing, or sound.

Yes. The model supports reference-guided video editing and controllable video extension. Clear instructions about what to preserve and what to change usually produce the most useful direction.

No. This is an independent creative interface that provides access to supported AI models. Seedance is a model developed by ByteDance Seed; this site is not affiliated with or endorsed by ByteDance.

Describe the subject and action first, then add setting, shot size, camera movement, lighting, pacing, visual style, and sound. When you upload references, state exactly what each file should control.