Google DeepMind has unveiled Gemini Robotics On-Device<\/strong>, a new AI model that enables robots to perform complex tasks locally without relying on an internet connection. The announcement marks a significant leap toward autonomous robotic systems<\/strong> capable of operating in real time using onboard computation.<\/p>\n\n\n\n Building on the cloud-based Gemini Robotics model introduced in March, this version allows developers to control robot behavior through natural language prompts and fine-tune responses to fit specific applications. Google says the new model matches or nearly matches the performance of its cloud-based predecessor<\/strong> and outperforms competing on-device models, although no direct comparisons were disclosed.<\/p>\n\n\n\n In a demonstration, Google showcased robots using the new model to unzip bags, fold clothes, and perform precise assembly tasks<\/strong> on industrial belts. Notably, a bi-arm Franka FR3 robot successfully executed tasks it had never encountered before, including handling unfamiliar objects and environments.<\/p>\n\n\n\n Though initially trained for ALOHA robots, the model adapted effectively to the Franka FR3 and Apptronik\u2019s Apollo humanoid robot, underscoring its versatility across hardware platforms<\/strong>.<\/p>\n\n\n\nDemonstrations show practical, real-world capabilities<\/h2>\n\n\n\n
New SDK empowers developers with fast training tools<\/h2>\n\n\n\n