Google DeepMind introduced Gemma 4 12B, a mid-sized, encoder-free multimodal model designed to run locally on consumer laptops with 16GB of RAM. The model supports native audio inputs, targets agentic multimodal tasks, and is positioned between the smaller E4B and the larger 26B Mixture of Experts model with a reduced memory footprint.

Google DeepMind unveiled Gemma 4 12B, a compact yet powerful multimodal AI model designed to run efficiently on laptops without separate encoders for audio and visual inputs. The model delivers near-top-tier performance with a smaller memory footprint, enabling advanced reasoning and agentic capabilities on consumer hardware. It marks a step forward in making high-performance AI more accessible for developers and everyday devices.