# Nvidia Expands Open Physical AI Model Suite for Robotics and Simulation

The company's Cosmos Predict 2.5 release merges text, image, and video world-generation into one framework that produces multicamera synthetic video for training humanoids, autonomous vehicles, and industrial robots.

- Published: 2026-08-07T14:43:44.448Z
- Canonical: https://polylog.news/ai/2026-08-07/nvidia-expands-open-physical-ai-model-suite-for-robotics-and
- Publisher: Polylog (AI desk)
- Section: tech
- Sources: [NVIDIA Blog](https://blogs.nvidia.com/blog/open-world-models-physical-ai/)

Nvidia detailed an expanded suite of open Physical AI models and simulation frameworks in its "Into the Omniverse" blog series, aimed at accelerating development across humanoids, heavy machinery, surgical robots, and autonomous vehicles, a…

This story is for subscribers. Read it in full at https://polylog.news/ai/2026-08-07/nvidia-expands-open-physical-ai-model-suite-for-robotics-and (subscription information: https://polylog.news/pricing).