← Back to GPN · All articles

NVIDIA's GR00T 2 Teaches Robots Dexterous Skills in Simulation

OFFICIAL Written by GPN · Nov 9, 2026, 4:00 PM ET · · Tech

Robot training is moving past walking and rolling. At this week's Conference on Robot Learning (CoRL) in Austin, Texas, NVIDIA introduced a set of open robotics technologies aimed at teaching robots tasks that need a finer touch — aligning gears, connecting cables and assembling small components — with simulation doing the heavy lifting.

The headline is GR00T 2, an open frontier world action model built on NVIDIA Cosmos and trained on diverse robot and human data, including 100,000 hours of NVIDIA open source human egocentric data. Out of the box it completed 40 of 50 evaluated tasks at least once with no task-specific demonstrations, and it reached 71% average success across 50 tasks with a single demonstration each, rising to 95% with five. It ranks among the top three models across the RoboCasa365, RoboLab, RoboDojo and MolmoSpaces benchmarks. Nexuni is testing it for robotic laundry handling, Foxconn is applying it to factory automation for server manufacturing, and Siemens is evaluating it across industrial use cases.

The training ground is Isaac Lab 3.0, the latest release of NVIDIA's open source robot-learning framework. It lets researchers use agents to set up training environments quickly, switch between NVIDIA PhysX and Newton physics solvers, and apply a new reinforcement post-training library to foundation models. Tighter integration with Newton, an open source physics engine, lets Isaac Lab model deformable objects such as cables and cloth, and in a vision-based dexterous manipulation benchmark it reached up to 2.5x the training throughput of Isaac Lab 2.3 with PhysX.

NVIDIA also introduced two new open benchmarks for measuring robot dexterity — DexBench, with 15 tasks across retail, logistics, manufacturing and hospitality scenes, and VABAR, which evaluates robots across six industrial tasks — plus a developer preview of Isaac Lab-Arena 0.4 for scaled policy evaluation. Two NVIDIA papers are among six finalists for the Best Paper Awards at CoRL.

Facts gathered from

• NVIDIA

More on this

Share