SoonMaker

Seedance 2.0 — Reference-Driven Video by ByteDance

Seedance 2.0 is ByteDance's frontier video model, built around reference-driven generation. Feed it reference images and it keeps the subject and style consistent through the shot, which is the hardest part of making AI video usable for anything beyond a single clip. It supports text-to-video and image-to-video with durations up to 15 seconds and native audio. Ideal when character or product consistency matters.

Video model · ByteDance

Seedance 2.0 — Reference-Driven Video by ByteDance

Frontier video that holds subject and style across the shot using multiple reference images.

Try Seedance 2.0 FreeSee examples

Try a prompt

A model walking through a neon-lit arcade, camera tracking alongsideA chef finishing a dish, steam rising, warm kitchen lightA vintage car pulling away on a coastal road at golden hour

Key features

6 highlights

Reference-driven subject consistency

Multiple reference images supported

Native audio generation

Durations up to 15 seconds

Text-to-video and image-to-video

Strong style retention across the shot

Specifications

Capabilities &
supported configs

  • Resolution480p, 720p, 1080p
  • Duration4–15 seconds
  • Aspect ratios21:9, 16:9, 4:3, 1:1, 3:4, 9:16
  • AudioAutomatic sound generation
  • Start frameNot supported
  • End frameNot supported

FAQ

Frequently asked

Ready to try Seedance 2.0?

Start generating with free credits. No credit card required.

Try Seedance 2.0 Free