Local LLM model fit

Can my GPU run Gemma 4 31B?

Gemma 4 31B is a 31B Gemma model. This page estimates Q4 VRAM fit, Ollama command, context planning, and fallback choices for common local AI GPUs.

Check Gemma 4 31B in the calculator

Q4 runtime estimate20 GB
Ollama commandollama run gemma4:31b
Recommended GPURTX 3090/4090 24GB minimum, 32GB+ preferred

Best use

High-quality multimodal reasoning, coding assistants, and local-first agent workflows. Weakness: Single 24GB GPUs have limited headroom for long context.

GPU fit table

HardwareExamplesClean capacityQ4 needStatusCalculator
6 GB VRAM entry GPUGTX 1660, RTX 2060 6GB4.50 GB clean VRAM20 GBRAM offloadOpen calculator
8 GB VRAM mainstream GPURTX 3060 Ti, RTX 4060, RTX 30706.50 GB clean VRAM20 GBRAM offloadOpen calculator
10 GB VRAM older high-end GPURTX 3080 10GB8.50 GB clean VRAM20 GBRAM offloadOpen calculator
12 GB VRAM local agent GPURTX 3060 12GB, RTX 407010.5 GB clean VRAM20 GBRAM offloadOpen calculator
16 GB VRAM creator GPURTX 4060 Ti 16GB, RTX 408014.5 GB clean VRAM20 GBRAM offloadOpen calculator
24 GB VRAM homelab workstationRTX 3090, RTX 409022.5 GB clean VRAM20 GBRuns locallyOpen calculator
48 GB VRAM workstationRTX A6000, L40S 48GB46.5 GB clean VRAM20 GBRuns locallyOpen calculator
Apple Silicon 32 GB unified memoryM2 Max 32GB, M3 Max 36GB26 GB unified20 GBRuns locallyOpen calculator

Quantization memory estimate on a 12GB GPU preset

QuantizationEstimated memoryUse case
Q4 / 4-bit20 GBDefault local inference balance
Q5 / 5-bit25 GBBetter quality, more VRAM
Q8 / 8-bit40 GBHigh quality, much more VRAM
FP16 / 16-bit80 GBMostly workstation/server use