ByteDance Launches Seedance 2.5: Generating 30-Second Audio-Video Clips in One Pass
petewoodbridge · x · 2026-08-07
ByteDance's Seed team has introduced Seedance 2.5, its new-generation video creation model. Advancing in long-form storytelling, multimodal reference, and editing, key highlights include:
- One-Pass Generation: Capable of generating high-quality, 30-second audio-video clips in a single pass, bringing a complete story to life in one take.
- Multimodal References: Allows users to input up to 30 images, 10 video clips, and 10 audio clips as reference materials in a single pass.
- Timestamp-Level Control: Offers targeted editing of audio and video content, significantly improving efficiency and controllability.
The API service is now available on BytePlus ModelArk, and users can also try it on Dreamina, CapCut, and other platforms.
Related event: Seedance 2.5 Global Launch: 30s Multi-Reference Video Generation(62 posts)→
More from Models
- Moonshot Joins Open-Weight Race as Kimi K3 Escapes Sandbox — Nunki08 · 2026-08-07
- Gemma 3 27B QAT Fidelity Regression: Why Attention Layers Need More Than Q4_0 — dampflokfreund · 2026-08-07
- SemiAnalysis: OpenAI Overcomes Pre-training Bottleneck, Develops New Model 'Doug' — ilkamoi · 2026-08-07
- Users Complain Claude Opus 5 Shifted from Sycophantic to Condescending — TheTuringPost · 2026-08-07
- Zhipu's GLM Coding Plans Get More Expensive as Chinese Models Shift Pricing — bookwormengr · 2026-08-07
- Ant's Ling 3.0 Tiny Activates Only 1.3B of 7.9B Params for Agents — truecakesnake · 2026-08-07