Gemini Robotics

Published

Mar 12, 2025

Gemini Robotics is an AI model developed by Google DeepMind (United States), first published in March 2025. It works in the robotics, vision and speech domain, on tasks such as instruction interpretation, robotic manipulation, speech recognition (asr) and 2 more.

Epoch AI has no training-compute estimate for this model. Training ran on Google TPU v4,Google TPU v5e.

Access: Unreleased. Its weights are not openly released. It is built on top of Gemini Robotics-ER.

Full record

Organization: Google DeepMind
Country of organization: United States
Domain: Robotics, Vision, Speech
Task: Instruction interpretation, Robotic manipulation, Speech recognition (ASR), Object recognition, Object detection
Training hardware: Google TPU v4, Google TPU v5e
Model accessibility: Unreleased
Open weights: No
Base model: Gemini Robotics-ER