Google DeepMind Unveils Gemini Robotics ER2 Model

July 30th, Google DeepMind launched the Gemini Robotics ER 2 model. It is currently their most powerful embodied reasoning model. The core focus is the robot's "high-level brain," responsible for understanding the real physical world, capable of complex multi-step task planning and arrangement, progress tracking and failure detection, invoking tools (such as Google search, custom functions), and commanding the underlying Vision-Language-Action (VLA) model to execute specific actions.

It forms a complete Gemini Robotics 2 series along with Gemini Robotics 2 (VLA Action Model) and On-Device 2 (on-device lightweight version).

Click the original article link below to join the Moving Observation Beating·Feishu AI News Channel, monitoring global AI hotspots and news 24/7.