1
Choose your input mode
Start from a text prompt for pure text-to-video, or upload one or two reference images when you already have a visual reference you want to animate.
CutFly Studio
Seedance 2.0
Reference Images (max 9)
Seedance 2.0 does not support uploading images with real human faces as references.
Click to upload
png, jpg, jpeg, webp (9 remaining)
Reference Videos (max 3, total 15s)
Reference Audios (max 3, total 15s)
0/5000
8s
4s15s
Generate with Audio
Generate audio alongside the video (increases credit cost)
Available Credits: 0
Cinematic Scene
1 / 2Generate AI Video with
Seedance 2.0
Seedance 2.0 is ByteDance's latest video generation model. Use it for text-to-video, image-to-video, dynamic camera control, and optional audio — all from a single prompt with up to 15 seconds of output.
Up to 15s
Clip length
720p
Max resolution
Audio
Optional generation
How it works
How to use Seedance 2.0
Four steps from idea to finished clip — text or image input, camera and audio control, then review
2
Write the prompt with scene and sound
Describe the subject, action, camera movement, and ambient sound in one prompt. Seedance 2.0 responds well to prompts that treat picture and audio as one unified scene.
3
Set duration, ratio, and audio
Choose between 4 and 15 seconds, pick the aspect ratio that matches your publishing channel, select 480p or 720p, and decide whether to include generated audio.
4
Generate and review the full result
Review motion quality, camera behavior, and audio alignment together. If any layer feels off, refine the whole prompt rather than adjusting only one element.
Ready to generate your first Seedance 2.0 clip?
Community Showcase
Real Examples from X
See what creators are building with Seedance 2.0 — real videos shared by the community on X.
Core capabilities
Why creators use Seedance 2.0
Longer clips, flexible input modes, dynamic camera control, and optional audio — in one workflow.
Text-to-video with no image required
Seedance 2.0 can generate a complete video clip from a text prompt alone. Use this mode when the full visual concept should come from language rather than a reference frame.
Image-to-video with one or two frames
Supply a first frame, a last frame, or both to anchor the visual direction. This gives you tight control over the scene's start point, end point, or both at once.
Dynamic camera control
Seedance 2.0 supports advanced camera movement as a first-class feature. Describe dolly, pan, tilt, or other moves in your prompt, and optionally lock the lens for stable reference shots.
Optional audio generation
Enable audio to have Seedance 2.0 generate sound alongside the visuals. This is useful for scenes where ambient sound, dialogue, or music timing matters — though audio is optional and increases the credit cost.
Clips up to 15 seconds
Seedance 2.0 supports output from 4 to 15 seconds. That range is long enough for product reveals, social hooks, explainer intros, and short narrative scenes without breaking into false long-form promises.
Multimodal reference inputs
Beyond image frames, Seedance 2.0 accepts up to 9 reference images, 3 reference videos, and 3 reference audio files to guide the generation. This makes it flexible for complex creative briefs.
Model workflow
Seedance 2.0 prompt ideas
Seedance 2.0 works best when prompts describe the subject, camera movement, and sound intention together. These examples show how to write prompts that use the model's full range.
Seedance 2.0 workflow on CutFly
- 1Choose text-to-video when the visual concept starts from a prompt, or image-to-video when you have a reference frame.
- 2Write one prompt that covers subject, camera movement, sound intention, and overall mood.
- 3Pick the aspect ratio for your target channel before generating — it shapes how the scene is composed.
- 4Use 480p for fast concept testing and 720p for cleaner final-quality clips.
- 5Enable audio when the scene needs ambient sound, dialogue, or music timing in the output.
Product reveal
Camera-led reveal with optional audio.
A sleek smartphone rotates on a dark reflective surface, slow 360-degree camera orbit, rim lighting, soft ambient hum, 720p, 8 seconds, 16:9.
Talking-head scene
Speech and camera movement working together.
A presenter speaks confidently to camera, gentle push-in, natural room tone, clean studio light, measured pacing, 720p, 8 seconds.
Cinematic nature shot
Best for landscape and environment clips.
A mountain ridge at golden hour, slow drone-like pan left to right, wind ambience, deep cinematic color grade, 16:9, 12 seconds.
Explore related model routes
These routes cover adjacent workflows when Seedance 2.0 is not the right fit for the job.
Kling 2.6
Explore Kling 2.6 if you want to compare another established video model route with a different motion and scene style.
Veo 3 model page
Use Veo 3 when native audio quality and audio-first scene design are the priority.
Compare AI models
See how Seedance 2.0 compares to Veo 3, Runway, and other models on CutFly.
Pricing and credits
Check the current credit packs before running high-volume Seedance 2.0 generations.
Use cases
Best use cases for Seedance 2.0
Choose Seedance 2.0 when longer clips, camera control, or multimodal references are part of the creative brief.
01
Short-form social content
Generate hooks, intros, and product moments up to 15 seconds for platforms like TikTok, Instagram Reels, and YouTube Shorts.
02
Brand and product videos
Create polished product reveals, ad concepts, and brand short-clips with precise camera control and optional audio.
03
Explainer and instructional clips
Use Seedance 2.0 for concise instructional moments where camera movement and a clear spoken line need to work together.
04
Storyboard and pre-visualization
Use it to visualize scene concepts, camera angles, and rough cuts before committing to a more polished final production workflow.
FAQ
Seedance 2.0 FAQ
Seedance 2.0 is a video generation model by ByteDance. It supports text-to-video, image-to-video with one or two input frames, dynamic camera control, and optional audio generation in a single workflow.