Google's new Gemma 4 12B puts AI agents directly on your laptop
google launched gemma 4 12b, a 16gb open model running locally on laptops, enabling advanced, private ai workflows without cloud dependence.
google introduced gemma 4 12b, a new open model designed to run locally on standard laptops with a minimum of 16gb ram, making advanced ai agents accessible without cloud reliance. the model utilizes a novel unified architecture that combines vision and native voice processing in a single system, supporting complex workflows and enhanced reasoning capabilities. additionally, google updated gemini 3.5 flash with computer use capabilities for agent automation and released nano banana 2 lite for faster image processing.