Physical AI / embodied world models
NVIDIA Cosmos
NVIDIA Cosmos is a family of world foundation models for physical AI. Cosmos focuses on multimodal reasoning, world simulation, synthetic data, action prediction, robotics, and autonomous systems rather than consumer 3D world creation.
Overview
| Status | Research |
|---|---|
| Access | Open weights / code |
| Released | 2025–2026 |
| Inputs | text, image, video, sound, actions |
| Outputs | video, synthetic data, action trajectories, world states |
| Best for | robotics, autonomous systems, synthetic data generation, physical AI reasoning, simulation workflows |
Why it matters
Cosmos is essential for the scientific and industrial side of world models. It connects the category to robotics, autonomous vehicles, simulation, and physical AI training data.
Roamscape use
Tracked as physical-AI world model infrastructure; useful for research pages and ecosystem mapping.
Strengths
- physical AI orientation
- robotics ecosystem integration
- multimodal reasoning and generation
- open model direction
Limitations
- not a consumer world generator
- requires technical infrastructure
- best suited to physical AI teams and developers