llmrun

Gemma 4 Models — Hardware Requirements

33 Gemma 4 models from Google and the community, from the smallest that runs in 1.3 GB of VRAM up to 32.7B parameters. Every row links to full quantization tables and GPU compatibility.

All Gemma 4 Models by Size

ModelParamsContext
Gemma 4 E2B IT Qat Mobile Transformers2.3B131K
Gemma 4 E4B IT Assistant4B131K
OFFELLIA Gemma 4 E4B 8B Claude 4.6 Opus Reasoning MTP4B—
Medgemma 4B IT4.3B—
Turkish Gemma 4B T1 Scout4.3B131K
Gemma 4 E2B IT Qat Q4 0 Unquantized5.1B131K
Gemma 4 E2B IT Qat Q4 0 Unquantized Heretic5.1B131K
Gemma 4 E2B IT5.1B131K
Gemma 4 E2B IT Uncensored5.1B131K
Gemma 4 E2B5.1B131K
Huihui Gemma 4 E2B IT Abliterated5.1B131K
Supergemma4 E4b Abliterated7.5B131K
Humanizer Gemma 4 E4b7.9B131K
Gemma 4 E4B IT Qat Q4 0 Unquantized7.9B131K
Gemma4 E4b Claims Comparison7.9B—
Gemma 4 E4B IT OBLITERATED8.0B131K
Gemma4 E4B MiniFantasy V18.0B131K
Gemma 4 E4B IT8.0B131K
Gemma 4 E4B IT Uncensored8.0B131K
Gemma 4 E4B8.0B131K
Gemma 4 E4B Luchador8.0B131K
Gemma 4 12B IT Heretic12.0B131K
Gemma 4 12B IT Qat Q4 0 Unquantized12.0B262K
Gemma 4 12B IT12.0B262K
Gemma 4 12B IT AEON Abliterated K4 BF1612.0B262K
Gemma 4 12B12.0B262K
Gemma 4 12B OBLITERATED12.0B131K
Gemma 4 12B IT Abliterated Uncensored12.0B131K
Gemma 4 12B Agentic Fable5 Composer2.5 v2 3.5x Tau212.0B262K
Gemma 4 12B IT Assistant12B262K
Gemma4 12B QAT Uncensored HauhauCS Balanced12B—
Gemma4 12B Mtp Assistant12B—
Gemma 4 19B19.0B262K
Gemma 4 26B A4B IT Uncensored25.8B262K
Gemma 4 26B A4B IT Uncensored Heretic25.8B262K
WaifuGemma4 26B A4b V125.8B262K
Gemma 4 26B A4B IT25.8B262K
Gemma 4 26B A4B IT Assistant26B262K
Gemma 4 26B A4B IT DFlash26B262K
Gemma 4 26B A4B IT Qat Q4 0 Unquantized26.5B262K
Gemma 4 26B A4B StyleTune v226.5B262K
Gemma 4 31B IT Qat Q4 0 Unquantized Assistant31B131K
Gemma 4 31B IT Speculator.eagle331B—
Gemma 4 31B IT DFlash31B262K
Gemma 4 31B IT Control Vectors31B—
Gemma 4 31B IT Scotoma 231.3B262K
Gemma 4 31B IT Uncensored Heretic31.3B262K
Gemma 4 31B IT Heretic31.3B262K
Gemma 4 31B IT31.3B262K
Gemma 4 31B IT Qat Q4 0 Unquantized32.7B262K
Gemma 4 31B IT Uncensored32.7B262K
Gemma 4 31B32.7B262K
Gemma 4 31B StyleTune32.7B262K
Gemma 4 Novelist Eclipse 31B32.7B262K
ExtGemma4 40 5B39.5B262K

How Gemma 4 Compares — Benchmark Rating

Gemma 4 26B A4B IT is the highest-rated Gemma 4 model with an overall benchmark rating of 50.2/100 — #34 among 78 open models. The top proprietary model, Claude Fable 5 (high), scores 87.8. Click a model to see its full benchmark breakdown.

Claude Fable 5 (high) · proprietary87.8
Claude Opus 4.7 · proprietary86.4
Claude Fable 5 Max · proprietary82.4
GPT 6 Astra (high) · proprietary82.4
GPT 6 Astra Max · proprietary82.4
Kimi K366.8
Qwen3.5 397B A17B66.1
GLM 565.9
DeepSeek V4.1 Flash65.8
DeepSeek R1 052863.1
Gemma 4 26B A4B IT50.2
Gemma 4 31B IT41.9
Composite of normalized public benchmark scores (methodology) · ■ Gemma 4 · ■ other models

Frequently Asked Questions

How much VRAM do I need to run a Gemma 4 model?
The smallest Gemma 4 model, Medgemma 4B IT, runs from 1.3 GB of VRAM at an aggressive quantization. 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?
46 of 55 Gemma 4 models fit in 16 GB of VRAM at some quantization, including Gemma 4 E4B IT, Gemma 4 31B IT, Gemma 4 26B A4B IT.
What is the most popular Gemma 4 model to run locally?
Gemma 4 E4B IT is the most downloaded Gemma 4 model in local-friendly quantized formats. It runs from 3.2 GB of VRAM.
How do Gemma 4 models score on benchmarks?
Gemma 4 26B A4B IT leads the family with an overall benchmark rating of 50.2/100, ranking #34 among 78 open models, while the top proprietary model, Claude Fable 5 (high), scores 87.8. See the comparison chart above for the full standings.
Gemma 4 Models — VRAM & Hardware Requirements | llmrun