Today for AI
HOT RADAR
opensourceHEAT 13.5°

AI CURATED HOT STORY · 11/7/2026

Video Rebirth Compresses 31B Video-Audio Diffusion Model to 14B with INT8

1 reports archived1 independent sourcesupdated 11/7/2026, 23:16:01
Synthesis & Latest Updates
1 Sources Cross-Validated

Video Rebirth released HyperFlow H3 Flash, a compressed few-step variant of MiniMax-H3 that shrinks a 31B joint video-audio diffusion model to 13.7B while supporting 4/6/8-step sampling without CFG. Weights, ComfyUI nodes and GGUF deployment are open-sourced; a 5.2s clip renders in ~92s on one RTX 5090 and 8.7s on 4×B200, with real-time inference on 4 Blackwell GPUs via SOL-H3.

LATEST/Video Rebirth released HyperFlow H3 Flash, a compressed few-step variant of MiniMax-H3 that shrinks a 31B joint video-audio diffusion model to 13.7B while supporting 4/6/8-step sampling without CFG. Weights, ComfyUI nodes and GGUF deployment are open-sourced; a 5.2s clip renders in ~92s on one RTX 5090 and 8.7s on 4×B200, with real-time inference on 4 Blackwell GPUs via SOL-H3.

Hourly heat curve

2 fully observed hours
Fewer than 3 fully observed hours; trend not plotted yet.

Coverage timeline

Total 1 reports · Latest first