Video Delta Net speeds up open video generation 75-90x: 14s video in 11s
xiuyu_l · x · 2026-09-03
Video Delta Net (VDN), a new open-source method, claims live text-to-video at faster-than-playback speed with near-lossless quality:
- Hybrid attention architecture enables real-time text-to-video generation
- Accelerates MiniMax-H3 by 75–90x: 14 seconds of 768p video generated in 11 seconds on 8× NVIDIA B200 GPUs
- Checkpoints, training/inference code, and a technical blog are all released
Related event: Video Delta Net Speeds Open-Source Video Generation 75–90x(2 posts)→
More from Multimodal
- jjk-explain turns any concept into a Jujutsu Kaisen-style explainer video with one Claude Code command — teortaxesTex · 2026-09-03
- WAN 3 Tops AI Video Editing Chart, Ranked #1 With Audio at 1189 Elo — koltregaskes · 2026-09-03
- Video Delta Net speeds up Minimax-H3 video generation 75-90x, faster than playback — xiuyu_l · 2026-09-03
- MiniMax H3 open-weight model generates character-drawing timelapses with cursor UI — WolframRvnwlf · 2026-09-03
- 'Reincarnated as a Vape': AI-generated short film leans into absurd premises — superfatbeagle · 2026-09-03
- Same-generation video keeps spatial layout: a location continuity test — Tokyo_Jab · 2026-09-03