Gemini 4 Argon for Robotics- Google Developers Blog
Oct 13, 2026Kanishka Rao

Summary
This post introduces Gemini 4 Argon, a physical reasoning model for robotics. It details its application in robot control (e.g., Spot mobile manipulation with MuJoCo practice, zero-shot control of robotic arms, and iterative steering of video world models for tasks like spelling with dexterous hands). It also covers Real2Sim capabilities, including reconstructing simulation twins from robot camera feeds and egocentric human video, and its use in physical understanding and research/self-improvement. Representative prompts and video demonstrations are provided for each area.
More from the Gemini-powered XR Development trail
- Jump to play: Building with Gemini & MediaPipe- Google Developers Blog
- Turn creative prompts into interactive XR experiences with Gemini- Google Developers Blog
- Veo 3 and Veo 3 Fast – new pricing, new configurations and better resolution- Google Developers Blog
- Build with Veo 3, now available in the Gemini API- Google Developers Blog
- Gemini 2.5 for robotics and embodied intelligence- Google Developers Blog