Gemma 4 is Google's open multimodal model family for building compact applications that understand text, images, and audio.
Coding & Dev

A lightweight open multimodal model family for local apps, with text, image, video understanding, long context, and multilingual support.
Best for: Developers who want an open Google model that can run on a workstation, laptop, or supported mobile and cloud setup.
Gemma 4 is Google's open multimodal model family for building compact applications that understand text, images, and audio.
Build local or edge AI experiences with an open multimodal model that understands text, images, and audio.
Meta open-weight multimodal models for text, image understanding, and long-context AI applications.