NVIDIA and Hugging Face Expand Robotics with GR00T N1.7 and Teleop

NVIDIA and Hugging Face are teaming up to enhance the open robotics ecosystem through the integration of NVIDIA’s Isaac GR00T N1.7 model and Isaac Teleop framework into LeRobot, Hugging Face’s open-source robotics library. These tools aim to streamline development by providing access to standardized frameworks, massive datasets, and advanced robotics workflows.

The GR00T N1.7 model, introduced in Early Access on April 17, 2026, is a cornerstone of NVIDIA’s physical AI strategy. As the first commercially viable open robot foundation model, it is trained on over 20,000 hours of human-centered data, enabling complex tasks like dexterous manipulation and multi-task learning. By integrating GR00T N1.7 into LeRobot, developers can adapt the model to new robot designs and tasks more efficiently.

Isaac Teleop further complements this by providing a framework for collecting high-quality human demonstration data. Developers can use it to create datasets in standardized formats, crucial for training and fine-tuning robot models. This integration supports NVIDIA’s larger goal of creating an end-to-end robotics development pipeline, from simulation to real-world deployment.

Expanding the Open Robotics Toolkit

Beyond GR00T N1.7 and Teleop, the collaboration introduces tools that streamline the robotics lifecycle:

• Isaac Lab-Arena: A simulation environment for prototyping and testing robot behaviors.
• NVIDIA Cosmos 3: An upcoming frontier model designed to supplement robotics data with simulated scenarios, useful where real-world data is scarce or expensive.
• Jetson Thor: Integration with LeRobot’s Reachy 2 robot for edge computing and deployment of vision-language-action (VLA) models.

The partnership connects NVIDIA’s 3 million robotics developers with Hugging Face’s 16 million AI builders, broadening access to state-of-the-art physical AI resources. Hugging Face cofounder Thomas Wolf emphasized the importance of open source in accelerating innovation, noting that these integrations enable developers to collaborate on training, evaluation, and deployment in a shared ecosystem.

Market and Industry Impact

For NVIDIA, this collaboration underscores its leadership in physical AI and robotics. The company’s broader Isaac platform—which includes GR00T, Teleop, and simulation tools—positions it as a one-stop shop for next-generation robot development. With a market cap of $4.77 trillion as of July 7, NVIDIA’s strategic push into robotics complements its dominance in GPU computing and generative AI.

Hugging Face’s role is equally pivotal. Known for its work in natural language processing, this partnership signals its intent to bring similar open-source principles to robotics, fostering a collaborative environment for researchers and developers.

Looking ahead, the planned integration of NVIDIA Cosmos 3 into LeRobot could further revolutionize the field by enabling data generation and policy simulation at scale. This would lower barriers for smaller teams, democratizing access to advanced robotics tools.

As robotics increasingly intersects with generative AI, NVIDIA and Hugging Face are laying the groundwork for an open, scalable, and collaborative future. Developers can explore the integration workflows today via the NVIDIA and Hugging Face platforms, with more updates expected as Cosmos 3 comes online.