Google DeepMind has published an update on Gemini Robotics 2, framing the system around the idea of bringing “whole body intelligence” to robots. The announcement positions the model as part of the company’s broader push to connect AI reasoning with physical machines.
The drafts broadly describe the advance as a move beyond narrow or segmented robot control toward more integrated robotic behavior. In that framing, whole-body intelligence means treating a robot’s physical form as a coordinated system rather than focusing only on a single limb, gripper or isolated capability.
The announcement underscores how robotics research is increasingly tied to general-purpose AI systems such as Gemini. While the source title does not provide implementation details, the core claim is clear: Gemini Robotics 2 is being presented as a step toward robots that can act with more coordinated, embodied intelligence.