Multimodal input
Reference text, images, video clips, and audio to guide style, motion, camera language, and sound.
Primary model
Seedance 2.0 for ecommerce videos, ads, and multimodal creation
Seedance 2.0 is the main model experience on this site. It supports text, image, audio, and video references, making it the best fit for product videos, ad variants, and controllable creative workflows.
Available in the generator
Best for
- Product photo to video
- Multimodal reference workflows
- Ad variations with controllable camera direction
- Complex motion and physically plausible scenes
Quick comparison
- Provider
- ByteDance Seed Team
- Inputs
- Text, image, video, audio references
- Common site workflow
- Text to Video, Image to Video, Reference to Video
- Best first page
- /product-video-generator
Better control
Use natural language to specify camera moves, lighting, product treatment, and scene progression.
Commercial workflow fit
The model is well suited for ecommerce, advertising, social content, and rapid creative testing.
Source used for this page
ByteDance Seedance 2.0 official model page
Seedance 2.0