Gemini Robotics-ER 1.6 Preview
gemini-robotics-er-1.6-preview
ProprietaryMultimodal
About
Gemini Robotics-ER 1.6 Preview is Google DeepMind's enhanced embodied reasoning model for physical AI and robotics tasks. Pointing and counting accuracy improved from 61% (ER 1.5) to 80%; multi-view success detection reaches 84% across simultaneous camera feeds. New instrument-reading capability enables robots to interpret complex gauges and sight glasses. Context window is 131,072 input / 65,536 output tokens. Available via Gemini API and Google AI Studio. Replaces gemini-robotics-er-1.5-preview, which was deprecated April 30, 2026.
Gemini Robotics-ER 1.6 Preview has a 128K-token context window.
Capabilities
VisionMultimodalReasoningFunction CallingTool UseStructured OutputsCode Execution
