GoogleGemini Robotics
Gemini Robotics-ER 2 Preview
Gemini Robotics-ER 2 Preview provides visual and spatial reasoning for robotics workflows.
Preview
What the release provides
It accepts text, image, video and audio input and generates text. The preview supports code and tool use for embodied reasoning across robot tasks.Gemini Robotics-ER 2 Preview official API documentation
How to use it well
Practical guidance. Connect its reasoning to a separate, validated action layer. Define coordinate systems, units and object references explicitly. A plausible plan or spatial answer is not proof that a movement is physically possible or that the robot has completed it.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Status | Preview |
| Access | API |
| Model identifiers | gemini-robotics-er-2-preview |
| Context window | 131,072 tokens API input limit; reserve output capacity where the service shares one window. Gemini Robotics-ER 2 Preview official API documentation |
| Maximum output | 65,536 tokens Published API output limit; request settings can be lower. Gemini Robotics-ER 2 Preview official API documentation |