At CES, NVIDIA unveils new open models and AI infrastructure for physical AI, partnering with industry leaders to debut AI-driven robots across all sectors. New models by NVIDIA aim to transform robots into reasoning generalists, enabling rapid learning of multiple tasks. NVIDIA’s technologies empower a global ecosystem to revolutionize industries with AI-driven robotics.

NVIDIA introduces new open-source simulation and compute frameworks to simplify robot development. Isaac Lab-Arena provides a collaborative system for large-scale robot evaluation and benchmarking, while OSMO offers a cloud-native orchestration framework. These frameworks aim to streamline complex pipelines and expedite the transition from research to real-world applications.

In collaboration with Hugging Face, NVIDIA integrates open-source Isaac and GR00T technologies into the LeRobot framework, accelerating open-source physical AI development. NVIDIA’s open models and datasets on Hugging Face lead downloads in the robotics community. This partnership aims to provide developers with streamlined access to integrated tools for efficient end-to-end development.

NVIDIA’s Jetson Thor meets the computing demands of humanoid robots with reasoning, showcased by leading developers at CES. NEURA Robotics, Richtech Robotics, AGIBOT, and LG Electronics unveil new humanoids integrated with Jetson Thor to enhance navigation and manipulation capabilities. The innovative module delivers 4x performance of the previous generation, ideal for energy-constrained autonomy.

NVIDIA extends robotics to the industrial edge with the new Jetson T4000 module, featuring the Blackwell architecture for autonomous machines and general robotics. The module offers high-performance AI computing and functional safety for $1,999 at 1,000-unit volume. Partners like AAEON, Advantech, Caterpillar, and more offer Thor-powered systems for edge AI, robotics, and embedded applications.

Read more at NVIDIA: NVIDIA Releases New Physical AI Models as Global Partners Unveil Next-Generation Robots