Cosmos 3: Omnimodal World Models for Physical AI research paper by NVIDIA, 2026

NVIDIA · Jun 1, 2026 · Multimodal and robotics · 90 citations · 142 upvotes

One model that reads and generates text, images, video, audio and robot actions for physical AI.

Summarised by hand from the abstract.

More from NVIDIA

Date	Paper	Topic	Citations
Jun 17, 2024	Nemotron-4 340B Technical ReportOpen models sized for one 8-GPU server, aligned with over 98% synthetic data.Foundation models · Jun 2024	Foundation models	138
Sep 17, 2019	Megatron-LM: Training Multi-Billion Parameter Language Models Using Model ParallelismMegatron-LM splits each Transformer layer across GPUs to train models with billions of parameters.Training and scaling · Sep 2019	Training and scaling	3,177

Topic

Date	Paper	Lab	Citations
Apr 5, 2023	Segment AnythingA promptable model and a dataset of over a billion masks that cut out any object in any image.Meta · Apr 2023Meta	Meta	15.7k
May 4, 2026	MolmoAct2: Action Reasoning Models for Real-world DeploymentA fully open robot action model, released with new datasets including 720 hours of two-arm teleoperation.Ai2 · May 2026Ai2	Ai2	44
Aug 6, 2026	Towards Physics of Multimodal Pretraining: Knowledge Flow, Modality Synergy, Early Unification, and RecipesControlled experiments on when text, image understanding and image generation help or compete when trained together.Meta · Aug 2026Meta	Meta	2

About this paper

Authors: Aditi, Niket Agarwal, Arslan Ali and 288 more
arXiv: 2606.02800 · PDF
Venue: arXiv.org
Citations: 90, 9 influential · Semantic Scholar
Upvotes: 142 · Hugging Face
Code: github.com/NVIDIA/cosmos
Lab: NVIDIA · on Companies · on Acquisitions · on Quarterly · on Paydays · on TechConf

Changes

Date	What changed
Sep 24, 2026	New paperAdded to the listSep 24, 2026

Sources: each lab's own papers and arXiv, with citation and upvote counts from Semantic Scholar and Hugging Face. One-line summaries are for orientation, not a substitute for the paper. Logos via logo.dev; trademarks belong to their owners.

What it shows

More from NVIDIA

Other multimodal and robotics papers

Changes

New papers by email