Google DeepMind Releases Gemini Robotics ER 2 for Embodied Reasoning
Google DeepMind has made Gemini Robotics ER 2, its embodied reasoning model for robotics, publicly available to developers via the Gemini API, Google AI Studio, and private preview on Gemini Enterprise Agent Platform. This model acts as a 'high-level brain' for robots, enabling them to understand the physical world, communicate with humans, plan multi-step tasks, and natively call tools like Google Search. It achieves 91.3% accuracy in moment-finding tasks with sub-second latency, suitable for real-world robot operation.
This release significantly advances the capabilities of physical AI, allowing robots to perform more complex tasks with improved understanding and real-time planning, bridging the gap between AI and robotics for practical applications.
Learn one new AI thing every day.
Daily Deck sends you seven plain-English cards like this every morning. Free.
Start free