1
Choose a Seedance model
Start with Seedance 2.0 for the best quality, or pick 2.0 Fast, 2.0 Mini, or 1.5 Pro from the model selector. The form updates its options to match.
AI SeedanceAI SeedanceText to video, image to video, and reference to video in one Seedance workspace
Pick a Seedance model, write the shot, and attach a start frame or up to 9 reference images. Choose 4–15 seconds, 480p to 4K, seven aspect ratios, and native audio — then download a clean MP4 with no watermark.
Create Seedance AI Videos in 4 Steps
Start with Seedance 2.0 for the best quality, or pick 2.0 Fast, 2.0 Mini, or 1.5 Pro from the model selector. The form updates its options to match.
Type a prompt, upload a start (and optional end) frame, or attach up to 9 reference images and tag them in the prompt.
Pick 4–15 seconds, 480p to 4K, an aspect ratio, and toggle Generate Audio. The credit cost updates before you submit.
Watch progress live, preview with sound, then download a watermark-free MP4 — or send it to Video Extend for the next shot.
Multimodal References, Native Audio, 4K Output, and Three Seedance 2.0 Tiers
Write the subject, action, environment, camera move, lighting, and sound. Seedance 2.0 turns a structured prompt into a physically grounded clip with multi-shot continuity.
Upload a start frame — and an optional end frame — to animate a photo, product shot, or storyboard panel. The model keeps identity, colors, and composition anchored to your image.
Attach up to 9 reference images and give each one a role in the prompt: @Image1 for the character, @Image2 for the outfit, @Image3 for the scene. Consistency holds across every cut.
Enable Generate Audio and Seedance renders dialogue, ambience, and effects with the picture — synchronized to motion and mouth movement, no separate audio pass.
Seedance 2.0 for final quality up to 4K, 2.0 Fast for quick 480p/720p drafts, 2.0 Mini for the lowest cost per second. Fast and Mini can also ground text prompts with web search.
Set any length from 4 to 15 seconds, render at 480p, 720p, 1080p, or 4K, and frame for 16:9, 9:16, 1:1, 4:3, 3:4, 21:9, or adaptive output.
How to Use Seedance 2.0 to Create AI Videos
The Seedance video generator runs text-to-video, image-to-video, and reference-to-video in one browser workflow. Use these steps to direct the visuals and audio, choose the output settings, and review the finished clip.
Use the playground at the top of this page — nothing to install. Sign in when you're ready to generate; Seedance 2.0 uses credits from your plan or credit pack.
Text to Video for a scene from scratch, Image to Video to animate a start frame (add an end frame for a defined finish), or Reference to Video with up to 9 tagged images.
Order it as subject → action → scene → camera → style → audio. Put dialogue in quotes. Refer to uploads as @Image1, @Image2 so each reference has one job.
Choose 4–15 seconds, 480p/720p/1080p/4K (4K on Seedance 2.0), a ratio for your platform, and whether to generate audio.
Preview the clip with sound, check faces, hands, text, and timing, then change one variable and re-run — or download and extend.
"Handheld push-in, subject turns to camera, fabric catches the wind" produces far more dynamic footage than a static description.
When a face, product, or outfit must hold, start from clear reference images and tag them in the prompt. Then verify consistency across the full clip.
With audio on, describe ambience, effects, and quoted dialogue separately from the visuals — Seedance times them to the action.
Test at 4–5 seconds on Seedance 2.0 Fast or Mini, lock the prompt, then render the final take on Seedance 2.0 at 1080p or 4K.
Seedance 2.0 for Every Creator
Vertical 9:16 clips for TikTok, Reels, and Shorts with motion, camera moves, and sound described in one prompt.
Product films, UGC-style ads, and campaign variations from a product photo and a few references — brand look locked across every frame.
Previz, moving storyboards, action sequences, and multi-shot narrative tests before committing to a full production.
Turn listing photos into rotating reveals, lifestyle scenes, and atmosphere shots, then verify product shape, color, and text.
Everything About Seedance 2.0 on AI Seedance
It's an online workspace for ByteDance's Seedance video models — Seedance 2.0, 2.0 Fast, 2.0 Mini, and 1.5 Pro. It supports text-to-video, image-to-video with start and end frames, reference-to-video with up to 9 images, native audio, and output from 480p to 4K.
Seedance 2.0 is the flagship: highest fidelity and the only tier with 1080p and 4K. Seedance 2.0 Fast returns 480p/720p clips faster for iteration. Seedance 2.0 Mini has the lowest cost per second for high-volume work. All three accept the same prompts, frames, and references.
Any length from 4 to 15 seconds. Seedance 2.0 renders at 480p, 720p, 1080p, or 4K; Fast and Mini at 480p or 720p; Seedance 1.5 Pro offers 4, 8, or 12 seconds up to 1080p.
Yes. Switch on Generate Audio and the model produces dialogue, ambient sound, and effects in the same pass as the video, synchronized to the motion. Describe the sound you want in the prompt and preview the full clip before downloading.
Yes. Choose Image to Video, upload a start frame, optionally add an end frame, and describe the motion, camera, and sound. For multi-character or multi-scene control, use Reference to Video with up to 9 tagged images.
Structure the prompt as subject → action → scene → camera → style → audio, put dialogue in quotes, and give every reference one explicit job (@Image1 = character, @Image2 = scene). Change one variable per run when refining.
Not in this workspace yet. Seedance 2.0 here accepts text, start/end frames, and up to 9 reference images. Video and audio references will be added when our provider exposes them.
No. Every clip you generate on AI Seedance downloads as a clean MP4 without a watermark.
Creating an account is free. Eligible new users receive 20 welcome credits, exactly enough for one 5-second Seedance 2.0 Mini video at 480p. The form shows and deducts the normal 20-credit cost; this is not a separate zero-credit trial. Further generations use available credits and are priced per second by resolution.
Commercial use depends on your plan terms and on the rights to your prompts and reference images. Clear any people, brands, music, or source assets before publishing, and avoid real faces or copyrighted characters without authorization.