Skip to content
Source
NVIDIA — Robotics· Ming-Yu Liu·· 59 days agoSignalEditorial score88

Into the Omniverse: How Open World Models Push the Frontier of Physical AI

Into the Omniverse: How Open World Models Push the Frontier of Physical AI

Summary

NVIDIA introduces Cosmos 3, an open-world model family for physical AI, combining vision reasoning, world generation, and action prediction. Available under the Linux Foundation’s OpenMDW 1.1 license, it enables post-training and adaptation for robotics, autonomous vehicles, and vision AI systems. The model ranks highly in benchmark evaluations and is part of NVIDIA’s broader physical AI stack.

Editorial context

NVIDIA introduces Cosmos 3, an open-world model family for physical AI, emphasizing its role in generating synthetic data, simulating future states, and enabling specialization for robotics, autonomous vehicles, and vision AI systems. The model is available under the Linux Foundation’s OpenMDW 1.1 license, supporting post-training and adaptation.

Evidence and limits

Published automatically after robotics and source-evidence checks; no manual editorial approval is recorded. Source assertions are not independently verified. Missing information remains not reported.

Environment:
Not reported
Control:
Not reported
Data origin:
Not reported
Source excerpts and review record

No manual editorial approval recorded.

Original source quotation: “Open models , which anyone can download, inspect, modify and run on their own infrastructure, are what make that possible.”

Source E1

Original source quotation: “NVIDIA Cosmos world foundation models are available under the Linux Foundation’s OpenMDW 1.1 license, enabling teams to post-train models on their own data and hardware.”

Source E2

Original source quotation: “Cosmos 3: The Frontier Model NVIDIA Cosmos 3 — a frontier open physical AI foundation omni-model built on a mixture-of-transformers architecture — combines vision reasoning, world generation and action prediction, letting developers use one model family to understand scenes, generate synthetic data, simulate future states and build specialized world action models .”

Source E3

Original source quotation: “Lightweight enough to run on edge GPUs, Cosmos 3 Edge can be deployed across NVIDIA RTX GPUs, NVIDIA DGX systems and NVIDIA Jetson, including Jetson Thor platforms.”

Source E4

Original source quotation: “Across benchmark evaluations, Cosmos 3 ranks No. 1 on Artificial Analysis for open weights text-to-image and image-to-video generation, on PAI-Bench for world generation and in the image-to-video category of Physics-IQ .”

Source E5

Original source quotation: “For robot policy, it ranks No. 1 on RoboLab . Cosmos 3 Super is also the highest-ranked open model on VANTAGE-Bench for vision understanding.”

Source E6

Source:NVIDIA — Robotics · blogs.nvidia.com