Gemma 4 Models — Hardware Requirements

35 Gemma 4 models from Google and the community, from the smallest that runs in 2.1 GB of VRAM at Q4 up to 32.7B parameters. Every row links to full quantization tables and GPU compatibility.

All Gemma 4 Models by Size

How Gemma 4 Compares — llmrun Composite Score

Gemma 4 31B IT is the highest-scoring Gemma 4 model with an llmrun Composite Score of 106.7 — #42 among 106 open models. The top proprietary model, Gemini 4 Carbon, scores 158.6. Click a model to see its full benchmark breakdown.

Frequently Asked Questions

How much VRAM do I need to run a Gemma 4 model?

The smallest Gemma 4 model, Gemma 4 E2B IT Qat Mobile Transformers, runs from 2.1 GB of VRAM at Q4_K_M (smaller low-bit quantizations start at 1.4 GB with a noticeable quality loss). Larger family members need proportionally more — see the table above for every model.

Which Gemma 4 models can I run on a 16 GB GPU?

37 of 62 Gemma 4 models fit in 16 GB of VRAM at Q4_K_M, including Gemma 4 E4B IT, Gemma 4 12B IT, Gemma 4 E2B IT.

What is the most popular Gemma 4 model to run locally?

Gemma 4 26B A4B IT is the most downloaded Gemma 4 model in local-friendly quantized formats. It runs from 16.5 GB of VRAM at Q4_K_M.

How do Gemma 4 models score on benchmarks?

Gemma 4 31B IT leads the family with an llmrun Composite Score of 106.7, ranking #42 among 106 open models, while the top proprietary model, Gemini 4 Carbon, scores 158.6. See the comparison chart above for the full standings.