Gemma 4: Byte for byte, the most capable open models
Summary
Google announces Gemma 4, the latest open-model family designed for advanced reasoning and agentic workflows. The four model sizes (E2B, E4B, 26B MoE, 31B Dense) offer on-device/offline capability with high efficiency, long context windows, multi-modal support (vision and audio), and broad language coverage; Gemma 4 is released under Apache 2.0 and supports integration with a wide ecosystem of tools and platforms.