Tech Desk 
Jun 25, 2025
Google DeepMind has officially launched a new Gemini Robotics artificial intelligence (AI) model that can run entirely on a local device.
Alphabet-owned tech giant on Tuesday, June 24, 2025, revealed that dubbed Gemini Robotics On-Device can make robots perform a wide range of tasks in real-world environments. The company said that since the AI model functions without the requirement to be connected to a data network.
Currently, the model is available to those who have signed up for its trusted tester programme.
Google's new robotics model runs on-device
In a blog post, a Senior Director and Head of Robotics at Google DeepMind, Carolina Parada stated that the new VLA model can be accessed via a Gemini Robotics software development kit (SDK) after signing up for its tester programme. The model can also be tested on the company's MuJoCo physics simulator.
The VLA model is designed for bi-arm robots and has minimal computational requirements. Despite that, the model allows for experimentation, and the company claims that it can adapt to new tasks with just 50 to 100 demonstrations.
Moreover, Gemini Robotics On-Device adheres to natural language instructions and can perform complex tasks such as unzipping bags or folding clothes.
Based on internal testing, the tech giant claims that the AI model “exhibits strong generalisation performance while running entirely locally.”
