Multimodal references
Guide a video with text, images, audio, or existing video clips.
SEEDANCE 2.0 · EARLY ACCESS PREVIEW
Seedance 2.0 AI Video Generator
Explore a faster way to turn text, images, audio, and video references into cinematic clips with ByteDance's Seedance 2.0 model.
Already have an account?
Text to videoImage to videoMulti-reference9:16 / 16:9
▶
✦ Two people walk along the coast at sunset, tracking shot…
9:16
SOCIAL READY
SOCIAL READY
● PREVIEW
00:05
00:05
01 / TRY SEEDANCE 2.0
Test your AI video prompt
This is a no-compute product preview. Your interaction helps us validate demand before we activate the paid Seedance API.
Examples:
Aspect ratio
Duration
SEEDANCE 2.0 ACCOUNT
Create your Seedance account.
Register now to save your place for real video generation and initial trial credits when they become available.
Checking your account…
02 / SEEDANCE 2.0 FEATURES
One multimodal workflow for AI video creation.
Audio-video generation
Seedance 2.0 uses a joint audio-video generation architecture.
Creator-ready formats
Plan vertical, landscape, or square clips for social workflows.
03 / HOW IT WORKS
From reference to video in three steps.
- 01Describe the shotSet the subject, action, setting, and camera movement.
- 02Add referencesUse an image now; audio and video references are planned.
- 03Generate with Seedance 2.0Real generation will open after demand validation.
04 / USE CASES
AI video for the formats people already watch.
@REC
Short-form content
Create concepts for TikTok, Instagram Reels, and YouTube Shorts.
◇NEW
Product videos
Turn a product image into motion concepts for ads and launches.
◉SCENE 04
Storyboards
Previsualize characters, shots, and worlds before production.