Gemma 3 4B

Gemma 3 4B is Google's 4-billion-parameter open-weight text model with a 131K token context window.

Strengths

Small parameter count makes it lightweight and efficient for resource-constrained inference while retaining a large context window for most tasks.

Best for

Local deployment, serverless inference, or any scenario where model size and latency matter more than raw reasoning capability.

Limitations

Smaller parameter count means less capability on complex reasoning or knowledge-intensive tasks compared to larger Gemma variants like Gemma 3 12B or 27B.