Gemini Robotics 1.5 enables agentic experiences, explains Google DeepMind - The Robot Report

Google DeepMind has introduced two advanced models, Gemini Robotics 1.5 and Gemini Robotics-ER 1.5, aimed at enhancing robotic capabilities toward artificial general intelligence (AGI) in physical environments. Gemini Robotics 1.5 is a vision-language-action (VLA) model that translates visual inputs and instructions into motor commands, enabling robots to perform complex tasks with transparent reasoning by thinking before acting. It also supports learning across different robot embodiments, accelerating skill acquisition. Gemini Robotics-ER 1.5, a vision-language model (VLM), excels in spatial understanding, reasoning about the physical world, planning multi-step missions, and natively calling digital tools such as Google Search. This model is accessible to developers via the Gemini API in Google AI Studio, while Gemini Robotics 1.5 is available to select partners. Together, these models form an agentic framework where Gemini Robotics-ER 1.5 functions as a high-level planner orchestrating robot activities, and Gemini Robotics