Cosmos 3: Omnimodal World Models for Physical AI
A unified world-model family that understands and generates language, images, video, audio, and actions for physical AI.
BibTeX citation
@article{cosmos32026,
title = {Cosmos 3: Omnimodal World Models for Physical AI},
author = {NVIDIA and others},
journal = {arXiv preprint arXiv:2606.02800},
year = {2026},
url = {https://arxiv.org/abs/2606.02800}
}