Google Gemma 3 4B IT

Gemma 3 is a multimodal AI model that processes vision-language inputs and generates text outputs, supporting context windows up to 128k tokens and over 140 languages; it enhances reasoning, mathematical, and chat capabilities with structured outputs and function calling, serving developers, enterprises, and AI builders seeking advanced multimodal understanding.