Gemma 4: Byte for byte, the most capable open models
Gemini
Google introduces Gemma 4, a high-performance, open-model family designed for advanced reasoning, agentic workflows, and efficient deployment across diverse hardware.Summary
Google has launched Gemma 4, its most intelligent open-model family to date, built on Gemini 3 technology and released under an Apache 2.0 license. The lineup includes four sizes—E2B, E4B, 26B MoE, and 31B Dense—optimized for tasks ranging from mobile edge computing to high-end workstation reasoning. These models feature advanced capabilities such as multi-step planning, native function-calling for agentic workflows, multimodal processing, and support for over 140 languages, providing developers with frontier-level performance and significant hardware efficiency.
(Source：Gemini)