ByteDance releases Seedance 2.5 with 30-second one-pass video generation and multimodal referencing
AI video model releases
ByteDance releases Seedance 2.5 with 30-second one-pass video generation and multimodal referencing
ByteDance's Seed team launched Seedance 2.5 on July 31, 2026, doubling single-pass generation length from 15 to 30 seconds with support for multi-round extensions to produce multi-minute videos.
The model accepts up to 30 images, 10 video clips, and 10 audio clips as reference material in a single pass, with strengthened clay render, motion, and creative reference capabilities.
Seedance 2.5 adds timestamp-level editing control for targeted audio and video edits, plus advanced features like green screen, camera perspective changes, and reference-based editing aimed at film and advertising use cases.
The model builds on Seedance 2.0's unified multimodal audio-video joint-generation architecture and focuses on foundational generation and reference-based generation.
Seedance 2.5 is rolling out now on Jimeng AI and Doubao Pro, with API access coming soon through BytePlus ModelArk.
cho.sh