News Ars Technica - All content Apr 2
Google announces Gemma 4 open AI models, switches to Apache 2.0 license
By Ryan Whitwam
88 score
AI Analysis
Building on yesterday's Social tease from Logan Kilpatrick, Google releases Gemma 4 open-weight models in four sizes (1B, 3B, 26B MoE, 31B Dense), switching from a custom license to Apache 2.0. The models are designed for local deployment, with smaller variants running on consumer hardware and larger ones on a single H100 GPU.
Google's Gemini AI models have improved by leaps and bounds over the past year, but you can only use Gemini on Google's terms. The company's Gemma open-weight models have provided more freedom, but Gemma 3, which launched over a year ago, is getting a bit long in the tooth. Starting today, developers can start working with Gemma 4, which comes in four sizes optimized for local usage. Google has also acknowledged developer frustrations with AI licensing, so it's dumping the custom Gemma license.
Open Source ModelsModel ReleasesGoogle AI
Source details
Google's Gemini AI models have improved by leaps and bounds over the past year, but you can only use Gemini on Google's terms. The company's Gemma open-weight models have provided more freedom, but Gemma 3, which launched over a year ago, is getting a bit long in the tooth. Starting today, developers can start working with Gemma 4, which comes in four sizes optimized for local usage. Google has also acknowledged developer frustrations with AI licensing, so it's dumping the custom Gemma license.
Open Source ModelsModel ReleasesGoogle AI