Examples
Seedance Video Generator
Multimodal AI video creation with director-level control
Seedance brings text, image, audio, and video inputs into one multimodal video creation workflow. Create, edit, and extend videos with stable motion, rich references, and precise camera control.
Unified multimodal creation from prompt to final cut
Start with text alone or combine images, videos, and audio references in a single workflow. Seedance helps teams move from concept to polished video faster while keeping visual intent, structure, and timing under control.
Immersive audio-video generation with stable motion
Built on a unified audio-video joint generation architecture, Seedance delivers stronger motion stability, more coherent scene transitions, and richer audiovisual alignment for ads, storytelling, and branded content.
Create, edit, and extend with cinematic precision
Go beyond basic text-to-video. Use Seedance to edit existing footage, extend clips forward or backward, and guide lighting, shadow, style, performance, and camera movement with a more director-like level of control.
Powerful Features
Multimodal Reference Control
Use text, images, videos, and audio together to guide subjects, style, framing, motion, music, and voice in a single generation task.
Audio-Video Joint Generation
Generate more immersive results with audiovisual coordination, including sound-aware scenes and support for audio-driven creative workflows.
Video Editing and Extension
Replace objects, refine scenes, repair regions, or extend clips into longer sequences while preserving continuity and creative intent.
Director-Level Camera Control
Shape movement, perspective, lighting, and shot rhythm with more precise control over camera language and cinematic presentation.
Create Professional Videos in 3 Simple Steps
Describe the scene and add references
Write a clear prompt, then add optional images, videos, or audio references to guide character appearance, framing, action, music, or voice.
Choose your mode and output settings
Generate from text, use first-frame or first-and-last-frame input, edit an existing clip, or extend a video. Then adjust ratio, duration, and other output settings to match your goal.
Generate, review, and iterate
Create the video, review the motion and audiovisual details, then refine your references or prompt until the result matches your creative direction.
Frequently Asked Questions
Everything you need to know about Seedance
What is Seedance?
Seedance is ByteDance's multimodal AI video creation model family. It supports text, image, audio, and video inputs, and can be used for video generation, video editing, and video extension in one workflow. Seedance 2.0 is one of its newer capability versions.
What inputs does Seedance support?
Seedance supports text prompts, reference images, reference videos, and audio inputs. You can use them separately or combine them for richer multimodal control.
Can Seedance edit and extend videos?
Yes. In addition to generating new videos, Seedance supports editing existing video content, replacing or refining elements, and extending clips into longer, more coherent sequences.
Does Seedance support audio-aware generation?
Yes. Seedance uses audio-video joint generation and can work with audio references to better align sound, rhythm, voice, and scene progression in supported workflows.
What output settings are available?
Available output settings vary by Seedance model. For example, Seedance 2.0 supports multiple aspect ratios including 21:9, 16:9, 4:3, 1:1, 3:4, and 9:16, with video durations from 4 to 15 seconds and output resolutions up to 720p depending on the selected model mode.
What makes Seedance different from basic video generators?
Its main strengths are unified multimodal input, audio-video generation, strong reference following, and built-in support for editing and extending video. Newer versions including Seedance 2.0 further strengthen these capabilities for more controllable, production-ready results.