Gemma 3 4B

What this model is

The architecture is dense. Every parameter is part of the active path, so memory and speed scale with the full model rather than a smaller expert subset.

Memory and quantization

Decimal GB, rounded from the Ollama tag page. Runtime memory still needs KV cache and overhead.

Estimate formula: published weight files + 1.1409 GB per 8K tokens × context / 8K + 1.5 GB runtime overhead. Sizes are decimal GB, not GiB. At 8K context, the smallest listed build is about 6.0 GB before OS and application reserve.

Conservative FP16 estimate from the Gemma 3 attention configuration; local sliding-window layers are capped in practice.

KV method source: https://ollama.com/library/gemma3

Speed heuristic

Speed depends on quantization, backend, and hardware; no single measured value is reported

Active path: 4B parameters. This is a relative speed signal only; it is not a tokens/sec benchmark.

Runtimes

• Ollama
• llama.cpp

ollama run gemma3:4b

License restrictions

Gemma Terms of Use and the Gemma Prohibited Use Policy apply.

This is a plain-language summary, not legal advice. Read the linked license before commercial deployment or redistribution.

Known issues

• Use is subject to the Gemma Terms of Use

Sources

• https://ollama.com/library/gemma3
• https://huggingface.co/google/gemma-3-4b-it

This page is an independent compatibility guide, not official documentation. Verify the primary source before downloading or deploying a model.