NVIDIA releases Cosmos 3 Edge: a 4B-param open world model for real-time robot reasoning and action on edge devices

NVIDIA 发布 Cosmos 3 Edge：4B 参数开源世界模型，为机器人及边缘 AI 提供实时推理与动作生成

NVIDIA open-sourced Cosmos 3 Edge on Hugging Face, a 4B-parameter world model that unifies scene understanding and action generation. It runs real-time at 15 Hz on Jetson Thor, producing 32 robot actions per inference. It ranks #1 on VANTAGE-Bench for vision analytics and sets a new SOTA for robot policy learning among 4B models. The architecture uses two transformer towers—autoregressive for reasoning, diffusion for prediction—with shared attention layers. The post doesn't disclose exact latency figures, only 'real-time inference,' so real-world performance will depend on the specific hardware and task.

Why it matters: NVIDIA open-sourced a 4B world model that runs real-time on Jetson Thor and directly outputs robot actions — size and practicality both hit the mark. Score held back from higher because it's just released, with no third-party benchmarks or cross-platform generalization results...