Google releases Gemini Robotics ER 2
Google DeepMind releases Gemini Robotics ER 2, an embodied-reasoning model for robots with text, image, video and audio input, continuous-video understanding, task orchestration, tool use and multi-robot collaboration.
What happened
Event details
Released on July 30, 2026, Gemini Robotics ER 2 serves as a high-level reasoning system for robots. Built on Gemini 3.5 Flash, it supports up to 128K context and text, image, video and audio input. It provides spatial and temporal reasoning, continuous-video progress tracking, tool orchestration and multi-robot collaboration, and is available to developers through the Gemini API and Google AI Studio.
Assessment
Why it matters
Gemini Robotics ER 2 extends Gemini's multimodal reasoning into embodied intelligence with continuous-video self-correction, tool orchestration and multi-robot collaboration, and makes those capabilities available through an API.
Availability
Access notes
The model is publicly available to developers, but its API model ID is marked preview. Gemini Enterprise Agent Platform remains in private preview.