Google Launches Gemini Robotics 2 for Advanced Humanoid and Multi-Robot Control

Google has unveiled Gemini Robotics 2, an advanced physical AI platform aimed at enhancing robotic capabilities. This new system is designed to give robots, particularly humanoids, comprehensive full-body intelligence, allowing for sophisticated movements and interactions. It also introduces features for advanced dexterity, enabling robots to perform intricate tasks, and facilitates seamless collaboration among different types of robots. The launch includes three core models: Gemini Robotics 2 for vision-language-action control, Gemini Robotics ER 2 for real-world video understanding and complex planning, and On-Device 2 for rapid adaptation to new hardware. These innovations aim to move physical AI beyond simple tabletop tasks, making robots more genuinely useful in diverse environments by enabling them to reason, collaborate, and execute multi-step plans.