What is Seedance 2.0?

Seedance 2.0 is a multimodal AI video generation model developed by ByteDance's Seed team. It uses a dual-branch diffusion transformer to generate video and audio simultaneously in a single pass, supporting up to 4K resolution and 15-second clips.

When was Seedance 2.0 released globally?

Seedance 2.0 launched in China between February 9–12, 2026, and became globally available on April 15, 2026 via Dreamina and CapCut.

How does Seedance 2.0 compare to Sora 2, Kling 3.0, and Veo 3.1?

Seedance 2.0 ranked first in both text-to-video and image-to-video categories on the Artificial Analysis Video Arena, outperforming Kling 3.0, Veo 3.1, and Runway Gen-4.5. It also supports up to 9 image references per request, compared to just 1 for Sora 2 and Kling 3.0.

What is the @-reference system in Seedance 2.0?

The @-reference system is Seedance 2.0's most distinctive feature. It automatically extracts camera paths from uploaded videos and rhythm/beat data from audio references, enabling highly consistent and stylistically controlled video generation.

Does Seedance 2.0 generate audio automatically?

Yes. Audio generation is included by default in Seedance 2.0 at no additional cost. The model generates video and audio simultaneously in a single pass using its dual-branch diffusion transformer architecture.

Seedance 2.0 Guide: Features, Architecture & Prompts

Seedance 2.0 Complete Guide: Features, Strengths & Prompts

A complete guide to Seedance 2.0: ByteDance's multimodal AI video model — covering architecture, core features, and practical prompts all in one place.

Quick readCRAISEE · Jun 24, 2026reference

Try it on CRAISEE nowGo create

Seedance 2.0 Complete Guide: Features, Strengths & Prompts