Gemma 4: Byte by byte the most powerful open models

At the edge, our E2B and E4B models redefine on-device utility by prioritizing multimodal capabilities, low-latency processing, and seamless ecosystem integration over raw parameter count.

Powerful, accessible, open

To power the next generation of groundbreaking research and products, we specifically sized the Gemma 4 models to operate efficiently and fine-tune hardware – from the billions of Android devices worldwide to laptop GPUs to developer workstations and accelerators.

By using these highly optimized models, you can fine-tune Gemma 4 to achieve state-of-the-art performance on your specific tasks. We have already had incredible success with this approach; For example, INSAIT has developed a groundbreaking model for the Bulgarian native language (BgGPT) and we have collaborated with Yale University, among others, on the Cell2Sentence Scale to discover new avenues for cancer therapy.

This makes Gemma 4 our most powerful open model family to date:

• Extended reasoning: Ability to plan at multiple levels…