Seedance

What is Seedance?

Key features of Seedance

• Multimodal input

  Seedance supports text, image, audio, and video as native input modalities, allowing creators to use one source or combine several. This suits workflows that bring scripts, visual references, sound, or existing footage into one video concept.

• Synchronized audio and video generation

  The model uses a joint audio-video generation approach to create moving imagery together with a corresponding audio track. This is useful for concept clips and audiovisual experiments where motion and sound should be developed as one creative result.

• Video extension and editing

  Seedance supports extending and editing video content around an existing clip, giving creators a way to explore new duration and visual variations from prior material. It fits shot continuation, creative iteration, and reworking footage rather than only starting from scratch.

• High-resolution visual output

  Seedance is described as producing 2K visuals together with a matched audio track in one generation process. This is relevant for clearer concept previews and early quality review, while final delivery suitability still depends on the project’s requirements.

• Character and physical consistency

  The model is described as supporting consistency constraints for characters and physical relationships, which helps when shots need coherent identity, movement, or scene changes. Creators can use this for continuity experiments and then select or correct results during review.