Google announces Gemma 4 open AI models, switches to Apache 2.0 license

Google has released Gemma 4, a new family of four open-weight models optimized for local inference, spanning from mobile-focused 2B/4B models to a 26B Mixture of Experts and a 31B Dense variant. Crucially, Google is dropping its restrictive custom license in favor of Apache 2.0, addressing a major developer complaint. The MoE model activates only 3.8B parameters at inference time, offering strong tokens-per-second on consumer hardware.