MODEL REFERENCE / google

google/gemma-4-12B-it

Google's encoder-free unified Gemma 4 dense model (12B) with native text, image, and audio, plus thinking mode and tool-use protocol.

Encoder-free unified multimodal model with audio, structured thinking, and function calling — runs on a single 40 GB+ GPU

vLLMdense12B131,072 ctxvLLM 0.23.0+multimodaltext

02 / Configure

Configure serving

Choose engine, hardware and deployment options. The command updates as you go.

03 / Understand

Model notes and examples