NVIDIA announced the NVIDIA Isaac GR00T Reference Humanoid Robot, describing it as the first open humanoid robot reference design built on NVIDIA Jetson Thor and the NVIDIA Isaac GR00T development platform. The company said the platform is designed to help democratize humanoid robotics research by combining advanced hardware, onboard AI computing, and an open software stack without requiring researchers to rely on proprietary systems.
The reference design integrates a Unitree H2 Plus humanoid robot, Sharpa Wave five-finger tactile hands, NVIDIA Jetson AGX Thor onboard computing, and the Isaac GR00T software platform into a single development system. NVIDIA said the platform is intended to streamline humanoid robotics workflows, spanning hardware integration, data collection, simulation, training, evaluation, and deployment.
According to NVIDIA, the robot stands nearly six feet tall, weighs approximately 150 pounds, and offers 75 total degrees of freedom across its body and hands. The system includes stereo vision cameras, wrist-mounted cameras, an inertial measurement unit, voice interaction capabilities, and a battery capable of delivering roughly three hours of operation. The onboard Jetson AGX Thor T5000 compute module is powered by NVIDIA’s Blackwell architecture and delivers up to 2,070 FP4 teraflops of AI performance.
The company said the Isaac GR00T platform includes tools for robot teleoperation, foundation models for humanoid reasoning and learning, simulation and training software through Isaac Sim and Isaac Lab, ROS middleware, and deployment capabilities on Jetson Thor hardware. Researchers can use the complete stack or integrate individual components into existing robotics development workflows.
Several research institutions are expected to adopt the platform, including Ai2, ETH Zurich, Stanford Robotics Center, and UC San Diego Advanced Robotics and Controls Laboratory. NVIDIA Research will also use the reference design to advance future Isaac GR00T models, frameworks, and hardware.
The company added that the Isaac GR00T developer platform will support the Unitree G1 humanoid robot, extending the ecosystem to another platform commonly used by robotics researchers and developers.
NVIDIA said the Isaac GR00T Reference Humanoid Robot will be available from Unitree Robotics in late 2026. The company also plans to release the Isaac GR00T reference workflow for the Unitree G1 on GitHub and Hugging Face.
KEY QUOTES:
“Humanoid robots will bring physical AI to the world’s largest industries, opening a multitrillion-dollar economic opportunity. The NVIDIA Isaac GR00T Reference Humanoid Robot gives researchers a single, open platform to make breakthrough discoveries toward general-purpose physical intelligence.”
Jensen Huang, Founder and CEO, NVIDIA
“Robotics moves fastest when researchers can build on open platforms, share code and test ideas on real machines. The NVIDIA Isaac GR00T Reference Robot gives our students and collaborators an open humanoid reference design with dexterous hands, onboard AI compute and the NVIDIA Isaac GR00T development platform for creating, comparing and sharing robot behaviors on physical hardware.”
Steve Cousins, Executive Director, Stanford Robotics Center
“ETH Zurich’s robotics research aims to advance machines that can move, perceive and manipulate reliably in the real world. The NVIDIA Isaac GR00T reference design gives our teams a state-of-the-art humanoid platform for collecting data, testing algorithms and validating robot behaviors with the NVIDIA Isaac GR00T development platform.”
Marco Hutter, Professor, ETH Zurich Robotic Systems Lab
“To make progress toward general-purpose robots, researchers need platforms that are both capable and broadly accessible. A reference design lets more researchers participate in frontier humanoid research and move from ideas to experiments faster. This helps push the whole robotics research ecosystem forward.”
Deepak Pathak, Cofounder and CEO, Skild AI
“At Ai2, our mission is to accelerate robotics through open science. The NVIDIA Isaac GR00T Reference Robot, built on NVIDIA’s open technologies, provides our researchers with the hardware and software components necessary to continue our work in broadly competent robotics.”
Dieter Fox, Senior Research Director, Ai2 and Professor, University of Washington
“Advancing robotics research for real-world problems requires humanoids that can move, interact and manipulate with precision in dynamic environments. An integrated platform that connects robot hardware, data capture, policy learning and physical evaluation can help researchers accelerate loco-manipulation research and develop more useful real-world systems.”
Michael Yip, Professor, UC San Diego and Director, Advanced Robotics and Controls Laboratory
