Menu

Seedance 2.0

Cinematic short videos from text, images, and refs with native audio sync

Open Studio

Not animated images — video with real shot designFrom idea to finished short film: text, reference images, first-and-last frames, reference videos, and audio — all handled in a single Seedance 2.0 model

Cinematic camera control

Push-in, tracking, orbit, close-up, low angle, handheld — real camera language, not AI motion collage. Generated frames look like they were shot with intent

Multi-shot storytelling, not a single looping frame

One prompt can describe opening, action, and closing shots. Framing and pacing follow the story naturally — well suited for ads, brand films, and short narratives

Consistent subjects and brands

Characters, outfits, products, and scenes stay stable across cuts without distortion — ideal for commercial content that needs to reuse characters and brand visuals

Realistic motion and physics

Cloth, water, smoke, vehicles, and human motion behave more naturally — reducing the jitter, floating, and unrealistic movement common in AI video

Seedance 2.0 model specifications

Core generation capabilities for creators and developers

Model

Seedance 2.0

An AI video model for short-form generation, image to video, reference-driven clips, and multi-shot storytelling

Generation modes

Text to video / Image to video

Supports text-only prompts, first-frame guidance, first-and-last-frame guidance, reference images, and reference video input

Reference inputs

Text / images / video / audio

Alongside the prompt, you can upload images, video, or audio as reference inputs

Aspect ratios

Adaptive, 16:9, 9:16, and more

Covers landscape ads, vertical social clips, square posts, and wider cinematic framing

Resolution

480p / 720p / 1080p

Choose output clarity based on preview speed, delivery quality, and asset use case

Duration

4-15 seconds

Fits product ads, social shorts, character motion, brand openers, and storyboard clips

Native audio

Supported

Can generate short video results with ambient sound, action audio, or musical atmosphere

Seedance 2.0 FAQ

What is Seedance 2.0?
Seedance 2.0 is an AI video generation model for creating short clips from text prompts and references. It is useful for product ads, social videos, character animation, brand openers, sports action, food videos, and cinematic storyboards
Can I create both text-to-video and image-to-video?
Yes. You can start from a text prompt only, or upload reference images to guide the first frame, product, character, scene, or visual style
What references can I use?
Seedance 2.0 supports six reference inputs: text-only, first frame, first-and-last frame, reference image (character / product / style), reference video (motion or composition), and reference audio (rhythm or dialogue). They can be used alone or combined
Does Seedance 2.0 generate audio?
Yes. You can enable native audio when you want speech, sound effects, ambience, or music direction to be generated with the video
Which ratios and durations are available?
Durations 4-15 seconds, resolutions 480p / 720p / 1080p, ratios 16:9, 9:16, 1:1, and 21:9. Available options depend on the selected profile and current workspace settings
When should I choose Seedance 2.0 instead of an image model?
Choose Seedance 2.0 when you need camera movement, product handling, human action, character motion, audio, or short-form video. Choose an image model for static posters, packaging, UI, or infographics
How does Seedance 2.0 compare to Veo 3, Sora 2, and Kling?
Seedance 2.0 is positioned for similar short-form AI video use cases as Veo, Sora, and Kling. Its strengths are multi-shot storytelling with consistent characters, synchronized native audio, reference-to-video (R2V) control, and 1080p output up to 15 seconds — making it especially useful for action sequences, branded ads, and multi-camera narratives where shot continuity matters
Which Seedance 2.0 settings should I choose?
Use lower resolution or shorter duration settings for draft exploration when available, then move to higher resolution, longer duration, or native audio settings for final-quality clips. Available settings depend on the current workspace profile
What is R2V (reference-to-video) in Seedance 2.0?
R2V — reference-to-video — lets you upload reference images or a reference video to guide motion, character identity, product appearance, scene composition, or visual style. Combined with text prompts, R2V keeps the generated clip close to your existing brand assets, characters, or storyboard frames
Can Seedance 2.0 keep the same character consistent across multiple shots?
Yes. Seedance 2.0 is designed for multi-shot storytelling with more consistent characters: identity, outfit, face, and proportions usually stay stable while the camera angle, background, crowd, lighting, and handheld motion change between shots
Does Seedance 2.0 support audio reference input?
Yes. In addition to text, image references, and video references, Seedance 2.0 can accept MP3 or WAV audio references. When native audio is enabled, the generated video can include both picture and sound
Can Seedance 2.0 extend or edit an existing video?
Partially. Uploading a reference video can guide motion style, scene composition, or visual direction, helping new clips stay coherent with existing footage. Seedance 2.0 is not a video editor — it does not support frame-by-frame editing, background keying, or precise subtitle layout. For final polish, use the generated result as a starting point in a dedicated editing tool
What kind of prompt works best for Seedance 2.0?
Specific prompts work best. Describe who the subject is, where the scene happens, what action occurs, how the camera moves, the lighting and style, the mood, and whether you need music, sound effects, or dialogue. For example: “At night in the rain, a runner crosses an alley, low-angle tracking shot, water splashes, fast breathing, cinematic lighting, 16:9”
How is Seedance 2.0 video generation priced?
Seedance 2.0 is billed by video tokens, which depend on resolution, duration, and aspect ratio. The lightest configuration (480p, 4 seconds) uses about 136 credits. Sign up and visit the pricing page to purchase a credit pack