Introducing Gemma 3: The most capable model you can run on a single GPU or TPU

• Gemma 3 is a collection of lightweight, state-of-the-art open models designed to run on devices.
• Gemma 3 outperforms other models in its size class, making it ideal for single-GPU or TPU applications.
• Gemma 3 supports over 140 languages and offers advanced text and visual reasoning capabilities.
• Gemma 3 includes a 128k-token context window and supports function calling for complex tasks.
• Gemma 3 comes with quantized versions for faster performance and reduced computational requirements.

Summaries were generated by Google AI. Generative AI is experimental.

The Gemma family of open models is foundational to our commitment to making useful AI technology accessible. Last month, we celebrated Gemma's first birthday, a milestone marked by incredible adoption — over 100 million downloads — and a vibrant community that has created more than 60,000 Gemma variants. This Gemmaverse continues to inspire us.

Today, we're introducing Gemma 3, a collection of lightweight, state-of-the-art open models ...