ByteDance Unveils Seedance 2.0: AI Video Generation Model Integrated into Doubao Assistant

ByteDance’s Seed team has rolled out Seedance 2.0, a cutting-edge AI video generation model now embedded in the company’s Doubao AI assistant. Available on Doubao’s mobile app, desktop, and web versions, Seedance 2.0 initially launched on the Jimeng platform before expanding to other ByteDance products like Doubao and Xiao Yunque.

Key features of Seedance 2.0 include multi-modal input (supporting up to 12 files: 9 images + 3 video/audio clips) for granular creative control; real-time audio-visual synchronization (millisecond-level) with accurate lip-sync across languages, action-matched sound effects, and music-driven camera movements; cinematic multi-lens storytelling with consistent characters and style; and video avatars via user identity verification.

To access Seedance 2.0: in the Doubao app, navigate to AI Creation > Seedance 2.0; on desktop/web, go to AI Creation > Video > Seedance 2.0 Fast. Users enter prompts, optionally upload reference images (excluding real people), select 5/10-second lengths, and wait for generation (peak hours may take hours).

Usage parameters: a 10-second video uses 2 credits (daily limit ~10 credits =20 points); 4-15 second videos require a paid subscription.

Sources

• https://www.yjpoo.com/site/7269.html