Text, image, video, and audio input
Describe what is new in text, then use references for details that are easier to show: a person, place, movement, camera path, performance, or soundtrack.
Seedance 2.0 Multimodal Video Generator
Create a polished 4–15 second shot from text and image, video, or audio references. Direct the performance, camera, continuity, and native sound, with resolution options from 480p to 4K.
209
Seedance 2.0 multimodal creation
A balanced production model for picture, motion, and sound
Seedance 2.0 accepts text, image, video, and audio input. Give each reference a defined role—identity, location, movement, camera, voice, or rhythm—so the model can follow a production brief instead of guessing how the files relate.
Use each reference for one clear purpose and make every shot change visible.
Native audio-video generation
Create dialogue, ambience, effects, and music with the image sequence so sound timing can be directed as part of the scene.
4–15 seconds from 480p to 4K
Use lower resolutions for working versions and select 1080p or 4K when the approved shot needs more delivery detail.
Multiple control paths
Start from text, animate a first frame, connect first and last frames, or use multimodal references according to the element that must stay stable.
Seedance 2.0 use cases
Use one model for movement, scale, and subtle performance
Seedance 2.0 is a strong general-purpose choice for short production shots that need more control than a prompt alone and more output detail than Mini or Fast.
00:10
01 / 03
Keep a fast subject readable through the camera move
00:10Direct an athletic or physical sequence while preserving the main subject, contact points, environment, and camera path.
How to use Seedance 2.0
Turn creative direction into observable instructions
Separate the brief into subject, action, camera, sound, and continuity. If an instruction cannot be seen or heard in the result, make it more concrete.
01
Define the change on screen
State who or what appears, the action that happens, and the final composition the shot should reach.
02
Assign every uploaded reference
Name which image, clip, or audio file controls identity, place, movement, lighting, voice, or rhythm.
03
Direct camera and sound together
Specify shot size, camera path, pace, dialogue, ambience, effects, and whether cuts or captions are allowed.
04
Review continuity before surface polish
Check faces, hands, props, spatial relationships, motion boundaries, and audio timing before judging texture or style.
Seedance 2.0 prompt guide
Write the prompt in the order a crew would execute it
Move from scene and subject to action, camera, sound, and continuity. Keep every instruction specific enough to verify in the generated video.
12-second Seedance 2.0 prompt example
SCENE — 16:9 cinematic realism at a warm night market. Preserve the chef from @Image 1 and the stall layout from @Video 1. ACTION — Follow the chef plating noodles, cut to a macro view of steam, then reveal the shared table. CAMERA — Smooth shoulder-height tracking, one clean cut, slow final pull-back. AUDIO — Natural cooking sounds and quiet conversation; no music or subtitles. CONTINUITY — Keep the face, apron, cookware, and light direction stable.
Establish scene and identity
Define the world, main subject, visual treatment, and the reference details that cannot drift.
Write the action in sequence
Use one readable action per beat and make the cause, order, and final state unambiguous.
Specify camera and timing
Name shot size, lens behavior, movement, cut points, and the moment when the final reveal should occur.
Describe sound and exclusions
State dialogue, ambience, effects, music, and anything that must not appear or be heard.
Seedance model comparison
Choose the Seedance model for this stage of production
Compare what changes the result in this workspace: resolution, duration, reference allowance, relative credits, and whether the priority is exploration, review, or final quality.
- Best for
- Narrative sequences, campaign films, detailed previs
- Visual target
- Highest fidelity and continuity
- Turnaround
- Quality-first
- Duration
- 4–30 seconds in this workspace
- Resolution
- 480p, 720p, or 1080p
- Reference limits
- Up to 50: 30 images, 10 videos, and 10 audio files
- Relative credits
- Highest
- Best for
- Prompt tests, social variants, product concepts
- Visual target
- Efficient everyday output
- Turnaround
- Fastest
- Duration
- 4–15 seconds
- Resolution
- 480p or 720p
- Reference limits
- Up to 9 images, 3 videos, and 3 audio files
- Relative credits
- Lowest
- Best for
- Polished action, character, campaign, and social shots
- Visual target
- High, balanced fidelity
- Turnaround
- Standard
- Duration
- 4–15 seconds
- Resolution
- 480p, 720p, 1080p, or 4K
- Reference limits
- Up to 9 images, 3 videos, and 3 audio files
- Relative credits
- High
- Best for
- Client previews, timely creative, motion tests
- Visual target
- Speed-quality balance
- Turnaround
- Faster
- Duration
- 4–15 seconds
- Resolution
- 480p or 720p
- Reference limits
- Up to 9 images, 3 videos, and 3 audio files
- Relative credits
- Medium
Seedance model selection
Choose Seedance 2.0 for the balanced production pass
Use the standard model when a short video needs high-resolution options, multimodal control, and native sound. Mini costs less, Fast shortens the feedback loop, and 2.5 extends the sequence.
Use Seedance 2.0 for polished short-form work
It fits cinematic shots, controlled action, campaign concepts, and social video that benefits from mixed references and output up to 4K.
Create with Seedance 2.0Use Mini or Fast before the final pass
Choose Mini for lower-credit exploration or Fast when reaching a creative review decision sooner is the main concern.
Compare Seedance 2.0 MiniChoose 2.5 for a longer sequence
Move to Seedance 2.5 when the story needs up to 30 seconds, a much larger reference set, or more detailed direction across multiple beats.
Seedance 2.0 FAQ
Seedance 2.0 multimodal-video questions
Answers about supported inputs, reference limits, video length, 4K, native audio, prompting, and model selection.
What can Seedance 2.0 generate?
Seedance 2.0 creates short multimodal video from text and optional image, video, or audio references. It supports directed motion, camera behavior, continuity, and native sound.
How long and how detailed can the video be?
This workspace supports 4–15 second output at 480p, 720p, 1080p, or 4K in several landscape, portrait, square, and wide aspect ratios.
How many image, video, and audio references can I use?
Supported reference workflows accept up to 9 images, 3 videos, and 3 audio files. Assign each file a clear purpose to reduce conflicting direction.
Does Seedance 2.0 generate synchronized audio?
Yes. It can generate dialogue, ambience, effects, and music with the visuals. Describe the role and timing of each sound, or turn audio off for a silent draft.
How do Standard, Fast, Mini, and 2.5 differ?
Standard offers the broadest resolution range up to 4K. Fast targets lower latency, Mini uses fewer credits, and 2.5 supports up to 30 seconds and a much larger reference set.
How should I structure a Seedance 2.0 prompt?
Start with format, scene, and subject. Add actions in order, camera movement, timing, and sound, then finish with continuity rules and clear exclusions.
Create with Seedance 2.0
Direct the picture, movement, and sound as one shot
Choose the format, assign each reference a purpose, and write the action in sequence to generate a controlled 4–15 second video.