Seedance 2.0

Add labels such as @Image1, @Video1 and @Audio1 to your prompt and link each label to an uploaded reference file. Images can guide composition and style, video can guide camera movement, and audio can inform rhythm. How closely the generated video follows these references depends on the prompt and inputs.

Seedance 2.0 supports text, images, video clips and audio tracks as inputs. The listed limits are up to 9 images (30 MB each), up to 3 video clips (2–15 seconds total, 50 MB each) and up to 3 audio tracks (15 seconds total, 15 MB each), with up to 12 files per request. Text prompt length is subject to the available interface and model limits. Check the options shown for the selected model before uploading.

Available resolution, frame rate, duration and aspect ratio depend on the selected model settings. Check the options shown for Seedance 2.0 before generating; the settings available in the interface determine the output.

Seedance 2.0 can generate audio alongside video. Prompts and audio references can guide dialogue, sound effects and background music. Lip-sync and audio-video synchronisation may vary with the language, prompt and inputs; an exact timing result is not guaranteed.

You can upload an existing video as a reference and ask the model to use its movement, camera path or pacing as a guide for a new video. You can also request changes to details such as clothing, actions or the scene. The result is a new generation, so it may not preserve the original structure exactly.

Seedance 2.0 supports video and audio references, up to 9 image references, an @-reference system for linking prompt elements to inputs, and video-to-video generation. Check the selected model's settings for available resolution, duration and speed options. These capabilities and performance can vary by model configuration; specific speed or output improvements are not guaranteed.

Generation uses account credits. The credit cost depends on the selected model and output options; check the displayed cost before generating. Sign in is required, and your account balance is checked before generation.

Seedance 2.0 may suit people creating videos who want to guide motion with reference clips, combine image, video and audio inputs, or explore generated dialogue, sound effects and music. It can be used for storyboards, narrated tutorials, branded content and other video concepts; results depend on the prompt, references and selected model settings.