Google DeepMind's new Gemini Robotics On-Device AI model Gemini Robotics On-Device AI from Google DeepMind allows robots to work completely offline, combining vision, language and action.

Gemini Robotics On-Device AI enables robots to operate without an internet connection, solving problems encountered in limited access environments. It combines vision, language and action, allowing robots to understand and act on natural language instructions. The model can quickly learn new tasks from few examples and adapts to different types of robots. While not as powerful as cloud-based versions, its reliability under offline conditions makes it promising for practical use. The technology is not yet commercially available, but it suggests a future in which home robots will be able to operate efficiently without the internet.