DeepMind Unveils Gemini Robotics-ER 1.6 Model

Google DeepMind has launched Gemini Robotics-ER 1.6, a new foundation AI model designed to enhance spatial reasoning, multiview understanding, and autonomy in physical robots. Key upgrades include precision object detection, relational logic, trajectory mapping, and instrument reading via agentic vision that combines visual reasoning with code execution. The model supports native tools including Google Search, vision-language-action models, and third-party functions. Boston Dynamics confirmed the model will power its Spot robot for fully autonomous real-world operation. Developers can access Gemini Robotics-ER 1.6 today through the Gemini API and Google AI Studio.