Gemma 4

Overview

Gemma 4 is the latest generation of open-source AI models from DeepMind built on Gemini 3 research. It delivers frontier-level capabilities in reasoning, multimodal understanding, and agentic workflows while remaining efficient enough to run on a wide range of devices from cloud servers to personal computers and edge hardware. It supports long context windows, function calling, and is available under a permissive open-source license for developers to use and fine-tune.

Key features

• Advanced reasoning
• Multimodal capabilities
• Agentic workflows
• Function calling support
• Efficient architecture
• Long context processing
• Open-source weights
• Multilingual support