LodeHQ
AI news for people who shipFree, by email

AI · haoailab.com · Saturday, 29 August 2026

Open-weight FastH3 renders 15s 768p video in 13s on one GPU

FastVideo, Nuva Lab, and NVIDIA FastGen distilled MiniMax's open-weight H3 into FastH3 Preview v1: a 4-step sparse checkpoint that generates 15-second 768p video with audio in under 13 seconds on a single Blackwell GPU, up to 14x faster than the base model. Full weights, a pre-extracted LoRA, and inference recipes are public, training code next. It is the first real post-training recipe for H3, and RTX, DGX Spark, and Apple MLX targets are already on the roadmap.

Read the source: haoailab.com ↗In the 29 August issue →