Models · MarkTechPost ·
Google DeepMind Releases Gemma 4 12B: An Encoder-Free Multimodal Model with Native audio that runs on a 16 GB laptop
Google DeepMind released Gemma 4 12B, an encoder-free multimodal model that takes vision and audio directly into the language model backbone. The model is available under Apache 2.0 and is said to run locally on a 16 GB laptop.