seedance-v2

To generate cinematic short-form video with native lip-synced audio, provide a text prompt describing the scene and speaking tone, then enable the generate_audio parameter to produce synchronized speech, SFX, or music within the 4–15 second output.

You can use image and video references to keep characters consistent by supplying up to 9 image URLs and 3 short video clips, allowing multimodal identity control while the scene evolves dynamically throughout the generated content.

The best way to create multi-language spokesperson dialogue ads is to provide a prompt specifying the speaking tone alongside an image reference for the character, utilizing native lip-synced audio for brand-consistent storytelling.

Seedance 2.0 Pro supports custom audio references for dialogue generation by accepting up to 3 short audio clips, enabling precise voice matching and synchronized speech output alongside the generated cinematic video.

Choose Seedance 2.0 Pro for cinematic short-form content requiring native lip-synced audio and multimodal identity control, while other models like HappyHorse 1.0 or Wan 2.7 may suit different video generation needs better.

Duration limits for cinematic video generation range from 4 to 15 seconds, with configurable aspect ratios and resolution settings, supporting up to 9 image inputs and 3 video inputs for reference-guided scene visualization.