Max Resolution
480p · 720p native output; 1080p availability depends on the serving platform/API tier
AI / Video Generation & Editing / ByteDance Seed
Seedance 2.0
Multimodal audio-video generation model supporting text, image, video and audio references.
Video GenerationMultimodal VideoNative Audio
Max Duration
4–15 seconds per generation
Generation Modes
Text-to-Video · Image-to-Video · Multimodal Reference-to-Video · Video Editing · Video Extension · Multi-Shot Audio-Video
Native Audio
Native joint audio-video generation with synchronized dual-channel audio
About Seedance 2.0
Seedance 2.0 uses a unified multimodal audio-video architecture with reference-driven creation, editing, extension and synchronized audio.
Specifications
Developer & Integration
1 specifications +
| API Access | Available through Volcano Engine Ark experience/API ecosystem |
Overview
4 specifications +
| Main Specialization | Unified multimodal audio-video generation, reference-driven creation, multi-shot storytelling and video editing |
| Target Audience | Creators, filmmakers, advertisers, e-commerce and game production teams |
| Best For | Cinematic short-form video, multi-shot storytelling, ads, product films, complex motion, reference-driven creation and synchronized audio-video scenes |
| Current Model | Seedance 2.0 |
Generation
7 specifications +
| Text-to-Video | Native text-to-video with prompt-driven camera planning, complex action and synchronized audio-video generation |
| Image-to-Video | Image-to-video plus multimodal reference generation for characters, scenes, composition and visual style |
| Video-to-Video | Video-reference, targeted editing and continuation workflows within the multimodal system |
| Reference-to-Video | Up to 9 images + 3 video clips + 3 audio clips + natural-language instructions |
| Video Extension | Stable prompt-controlled continuation and extension of existing video |
| Multi-Shot Generation |
