NVIDIA open-sources GR00T 2 robot foundation model with 100K hours of training data
DrJimFan · x · 2026-11-11
NVIDIA's GEAR lab introduced GR00T 2, its most advanced open "System One" robotics model, running entirely onboard, fully untethered, with every weight owned by the user.
Key capabilities:
- Long-horizon tasks: dexterous manipulation and whole-body control chained over 20+ steps and 5+ minutes, like chemistry lab experiments or end-to-end model airplane assembly
- X-Adapt: learning from easy data sources (human wearables, simulation) and deploying zero-shot to diverse real robots
- Unified Cognition: native integration with frontier agents for physical reasoning, error recovery, and safety monitoring
The open-source buffet includes weights, X-Adapt recipes (December), the agentic framework, simulation engine, annotation engine, hardware designs, and 100K hours (11 years) of training data — all available today.
Related event: NVIDIA Releases Open-Source Robot Foundation Model GR00T 2 at CoRL(18 posts)→
More from Embodied
- Perceptron open-sources EgoHome, a sub-second egocentric household-video benchmark — lukas_m_ziegler · 2026-11-11
- Anthropic touts Fable 5.2 robot control gains on RoboDojo; real-world box folding disappoints — _Stocko_ · 2026-11-11
- AI2 open-sources ToyBlocks to benchmark robotic world models by task-relevant fidelity — allen_ai · 2026-11-11
- OpenTac: open-source ecosystem turns any robot surface into a tactile skin — _krishna_murthy · 2026-11-11
- NVIDIA launches open Alpamayo 2 Edge for real-time in-vehicle AI reasoning — nvidia · 2026-11-11
- Stanford DroneHacks: 100 students fly GPS-denied PX4 drones on stereo cameras in 48 hours — mrpollo · 2026-11-10