Introducing Gemma 4 12B: a unified, encoder-free multimodal model

DeepMind Blog

Gemma 4 12B runs on laptops with 16GB of RAM, supports native audio and vision inputs, and is released under Apache 2.0. It delivers benchmark performance nearing Google's larger 26B MoE model while using less than half the memory.

GemmaGoogle

See 14 more sources

Bringing Gemma 4 12B to your Laptop: Unlocking Local, Agentic Workflows with Google AI Edge10 days ago

Gemma 4 12B: The Developer Guide10 days ago

yuxinlu1/gemma-4-12B-coder-fable5-composer2.5-v1-GGUF16 hours agoyuxinlu1

RT @AndreasPSteiner: Released last week, and already more than 4M downloads on HuggingFace alone 😊...11 hours agoHugging Face

RT @mervenoyann: Google dropped Gemma-4 12B, it's a beast 🔥 > unified: audio + image go...10 days agoHugging Face

RT @googlegemma: Meet Gemma 4 12B! A unified, encoder-free multimodal model designed to bring...10 days agoHugging Face

Google’s newly released open weights model, Gemma 4 12B, supports transcription but is far from the...8 days agoArtificial Analysis

Google's new open source Gemma 4 12B analyzes audio, video — and runs entirely locally on a typical 16GB enterprise laptop10 days agoCarl Franzen

Google Gemma 4 12B nearly matches 26B benchmarks — and runs on your laptop9 days agoMeredith Shubel

Google's new Gemma 4 12B model is designed to run on any laptop with 16GB of RAM10 days agoRyan Whitwam

google/gemma-4-12B · Hugging Face10 days agojacek2023Discuss

Google just dropped Gemma 4 12B on your laptop!!10 days agoNewMuffin3926Discuss

120 tok/s on 12GB VRAM with Gemma 4 12B QAT MTP7 days agojanvitosDiscuss

New Google Gemma 4 12B Claims Near-26B Performance - We Tested Both!10 days agogladkosDiscuss

DeepMind Blog

5 days ago