Galbot Upgrades Humanoid Robot with NVIDIA Jetson Thor

• Galbot has upgraded its G1 Premium robot with NVIDIA Jetson AGX Thor for better speed, energy efficiency, and control.
• The company is using this improvement to expand into more real-world environments, including pharmacies and factories.

Galbot is one of the first companies to use NVIDIA’s Jetson AGX Thor, upgrading its G1 Premium robot with faster, more efficient computing. The company says the move boosts speed, autonomy, and real-time reasoning. The enhanced G1 Premium was introduced at the World Robotics Conference, where it was named the “swiftest humanoid worker.” The robot also won the championship at the World Humanoid Robot Games.

Built on Jetson Thor, the G1 Premium has shown gains in motion fluidity, real-time reasoning, and complex task planning. The robot runs Galbot’s VLA models and uses a Sim2Real approach that trains systems on synthetic data with minimal real-world testing. Galbot also co-developed a simulation platform called OpenWBT_Isaac with Tsinghua University and Shanghai Qi Zhi Institute, using full-body teleoperation and NVIDIA GPUs to support advanced humanoid development.

Galbot’s G1 robots are already operating fully autonomously in more than 10 pharmacies in Beijing, with plans to expand to over 100 locations across China by year-end. The company also aims to grow its footprint in smart manufacturing through a partnership with Bosch Group. With Jetson Thor now in the loop, Galbot is accelerating its roadmap and using advanced compute to push toward broader commercial deployment.

🌀 Tom’s Take:

Jetson Thor may be new to the market, but early adopters like Galbot are already showing what this AI brain can do, and why it matters for real-world autonomy.

Source: PR Newswire / Galbot