Google DeepMind Gemini Robotics 2: Whole Body AI Control

Google DeepMind has released Gemini Robotics 2, an intelligence layer for its next generation of robots. This new release moves robot capabilities beyond basic table-top tasks. What's interesting is it allows for whole body control, five-finger dexterity, and multi-robot teamwork. It ships as three separate models with different access tiers. Most robots today are pre-programmed for narrow tasks and don't adapt to new environments. Gemini Robotics 2 aims to address these limitations. One of the models, Gemini Robotics 2, is a vision-language-action model that converts inputs into motor control. It can drive humanoids from feet to fingertips and handle dexterous manipulation. Another model, Gemini Robotics ER 2, acts as a high-level brain, planning multi-step tasks and communicating with humans. The third, Gemini Robotics On-Device 2, is optimized to run directly on the robot. This development means robots could soon perform more complex and adaptive tasks in the real world.