NVIDIA and Hugging Face add robotics tools to LeRobot

NVIDIA and Hugging Face are expanding LeRobot, Hugging Face’s open source robotics library, with new tools for physical AI development. The collaboration brings NVIDIA Isaac GR00T 1.7, an open reasoning vision language action model for humanoid robots, and NVIDIA Isaac Teleop into LeRobot, with NVIDIA Cosmos 3 planned soon.

The integrations are designed to give robotics developers a shared pipeline for collecting and standardizing data, training and fine-tuning robot foundation models, evaluating performance and deploying models through open workflows. Isaac Teleop supports human demonstration capture from external devices in interoperable formats, while Isaac GR00T 1.7 helps developers post-train and deploy models for new robot embodiments and tasks.

Cosmos 3 is expected to help developers generate and augment robotics data, simulate scenarios and support policy development when real-world data is limited or too expensive to collect. NVIDIA said the partnership connects its 3 million robotics developers with Hugging Face’s 16 million AI builders.

The effort builds on existing NVIDIA resources tied to LeRobot, including an open source physical AI dataset downloaded more than 15 million times, with more than 350,000 real and simulated trajectories and 57 million grasps. Other connected tools include Isaac Sim and Isaac Lab simulation frameworks, Isaac Lab-Arena in LeRobot Environment Hub and NVIDIA Jetson Thor integration with LeRobot’s Reachy 2.