Hansheng Chen, Shengqu Cai, Jiaming Song, Gordon Wetzstein, Xinchen Yan and Leonidas Guibas · Decoder-free few-step video
pi-H3
pi-H3 is a few-step distillation of MiniMax H3 that generates video with audio in 2 or 4 sampling steps. It predicts pixels directly at the final denoising step (DWD, Decoding Without a Decoder), so the video VAE decoder is skipped. On a single B200 it produces a 5-second 768p video with stereo audio in 3.8 s with 2 steps, and a 15-second video in 18.6 s. The authors report it as 2.4–2.7× faster than Sol-H3 at 2 steps and 1.2–1.4× faster at 4 steps, using their own measurements against Sol-H3's published figures. It combines pi-Flow v2, a data-free few-step distillation method, with DWD, which the authors call experimental on H3. It is an open-source research project: a Hugging Face demo is live, weights and inference code are due Nov 16–22 in LakonLab, and training code will come later in November.
Open the source ↗Similar Video & Animation releases
- PixVid: Back to Pixels, Lifting Latent Video Diffusion into Pixel Space
Yanke Song, Shuo Chen, Chris Rockwell, Ming-Hsuan Yang, Xun Huang et al. · Pixel-space video diffusion · Week 47, 2026
- MPEG-Diffusion: Mixed-Resolution Video Generation
Dennis Menn, Tianpei Gu, Guande He, Aryan V S, Yanke Song, Nan Liu, Desai Xie, Xun Huang, Chenfeng Xu and Diana Marculescu · Mixed-resolution video diffusion · Week 47, 2026
- Open-H3-Regenerate-2K
NVlabs (SANA / SoL-Engine team) · Open video refiner · Week 47, 2026
- Few-Step Streaming Video Generation via Condition Shifting
Tsinghua University (Deyuan Liu, Jungang Li, Kaiwen Zheng, Kai Jiang, Jintao Chen, Yaole Wang, Ming Lin, Tao Lin) · Streaming Video Distillation · Week 47, 2026
- NVIDIA Cosmos 3 Nano interactive world simulation checkpoints
NVIDIA · World Model Checkpoints · Week 46, 2026
- Genie 4
Google DeepMind · Interactive world model · Week 45, 2026
Published by Quanta AI Labs, an AI agency in India. Browse every week in AI News Weekly.