ByteDance Releases Seedance 2.5: Generates 30-Second Videos with Multi-Round Extension
字节跳动Seed · wechat · 2026-07-31
ByteDance has officially released its new video generation model, Seedance 2.5, focusing on enhanced long-form narrative, multimodal referencing, and precise editing capabilities.
- Long Video Generation: Single generation duration is increased to 30 seconds, with support for multi-round extensions. The model can organize multiple logically connected shots to tell a complete story, while effectively reducing the common "AI greasy feel."
- Multimodal Referencing: Supports up to 30 images, 10 videos, and 10 audio clips as inputs. Enhanced features like white-model and motion referencing allow for better reproduction of complex group scenes and physical lighting.
- Precise Editing: Supports timestamp-based directional audio/video editing, offering green screen replacement, viewpoint, and camera movement adjustments for professional film and ad scenarios.
The model is currently rolling out on Jimeng AI and Doubao Pro, with API services coming soon to Volcano Ark. ByteDance also noted its exploration into industrial applications like education, manufacturing, and autonomous driving synthetic data.
Related event: ByteDance Releases Seedance 2.5: Native 30s Video and Multimodal Control(60 posts)→
More from Multimodal
- Alibaba Opensources ClinFusion: A Medical Multimodal Foundation Model — aigclink · 2026-08-01
- MiniMax 3 passes Turing test for AI video: writes 'Hi' on chalkboard, now open source — Kyrannio · 2026-08-01
- Testing Doubao vs. Jimeng: $30 Gets You Only 3-6 AI Video Generations — oran_ge · 2026-08-01
- Experiment Shows Training AI Image Models at 512 Resolution + Upscaling Saves VRAM — More_Bid_2197 · 2026-08-01
- 8-Hour Debug of ComfyUI Black Images Uncovers PyTorch FP16 Overflow Bug — rtsitola · 2026-08-01
- MoGe-3 by Microsoft Sets SOTA on 9 Benchmarks for High-Fidelity 3D Geometry from a Single Image — RexDouglass · 2026-08-01