Now Available — Gemma 4 Released

Everything You Need to
Run Gemma 4

The complete resource for Google's latest open model. Tutorials, benchmarks, and deployment guides — from zero to production.

Maximum capability

Gemma 4 31B Dense

31B parameters

Research & Development

Efficient power

Gemma 4 26B MoE

26B parameters

High Performance
4 models in family

Edge optimized

Gemma 4 E4B

4B parameters

Consumer Hardware

Embedded AI

Gemma 4 E2B

2B parameters

Mobile & Edge
128K
Context Window
140+
Languages
Apache 2.0
License
3.8B
Active Params (MoE)
Capabilities

Built for the next generation

Gemma 4 sets new benchmarks for open models, delivering state-of-the-art performance across reasoning, code generation, and multimodal understanding.

Advanced Reasoning

Chain-of-thought reasoning for complex problem-solving

Agentic Workflows

Built-in tool use and function calling

Code Generation

Expert-level code across 140+ languages

Vision & Audio

Native multimodal understanding

Stay updated

Get notified when new models drop, tutorials publish, and the ecosystem grows.

No spam. Unsubscribe anytime.