Google DeepMind launches Gemini Robotics 2: whole-body AI for adaptable robots
DeepMind unveiled Gemini Robotics 2, a vision‑language‑action model that lets robots reason, plan, and control every limb—from walking to fine manipulation—across many bodies. The suite includes a powerful VLA, an embodied‑reasoning engine, and an on‑device version that can adapt to new hardware in hours, opening real‑world multi‑robot collaboration.