Google Launches Gemini Robotics ER 2 for Smarter Robots

Google has launched Gemini Robotics ER 2, its most advanced embodied reasoning model for robotics. The model acts as a high-level brain, enabling robots to chat with humans, plan multi-step tasks, track progress via continuous video feeds, and self-correct in real time. It also introduces multi-robot collaboration for complex shared workflows. Available now via the Gemini API and Google AI Studio, ER 2 outperforms its predecessor across tool orchestration, progress classification (57.4% accuracy), and moment-finding (91.3% accuracy) at four times the execution speed. A demo featuring Boston Dynamics Spot is available on GitHub.