Google introduces Gemma 4 12B, a unified, encoder-free open multimodal model that can run locally on devices with 16GB of VRAM or unified memory (Carl Franzen/VentureBeat)

Google has released Gemma 4 12B, an open-source, encoder-free multimodal model that processes audio and video. The model is designed to run entirely locally on standard laptops equipped with 16 GB of RAM or equivalent VRAM/unified memory, allowing it to operate without cloud resources.

本文综合以下全部报道写成，而非依据任何单一来源。