NVIDIA Open-Sources SANA-WM: 2.6B-Parameter World Model Generates 1-Minute 720p Video on a Single GPU — 371 Points on HN

NVIDIA Labs / arXiv·high signal

NVIDIA Labs released SANA-WM under Apache 2.0 — a 2.6B-parameter Diffusion Transformer that takes a single image and camera trajectory and synthesizes 60-second 720p video. A distilled variant runs on a single RTX 5090 with NVFP4 quantization, producing a 60-second clip in 34 seconds (2.1x real-time). Trained in ~18.5 days on 64 H100s with only 212,975 public video clips, achieving 36x higher throughput than LingBot-World (14B+14B) at comparable VBench scores.