CanMyPCRunAILocal AI compatibility
Check my PC

NVIDIA · 32GB VRAM

What AI models can a GeForce RTX 5090 run?

A GeForce RTX 5090 has 32GB of VRAM. Checked against 228 models from the Ollama library, 194 run fully on the GPU at 4K context and 10 more run with part of the model in system RAM.

194

run well

10

with trade-offs

228

models tested

Check my actual PC

Best AI models for a GeForce RTX 5090

Largest first — these fit entirely in 32GB of VRAM at 4K context, so the whole model is GPU-accelerated.

Models that run well on a GeForce RTX 5090
ModelBuildMemory neededVerdict
laguna-xs-2.1Ollama Librarylatest~27.4 GBGood fit
codeboogaOllama Library34B~27.3 GBGood fit
phind-codellamaMicrosoft34B~27.3 GBGood fit
qwqAlibaba Cloud32B~26.8 GBGood fit
command-rCohere35B~26.8 GBGood fit
olmo-3.1Allen Institute for AI32B~26.3 GBGood fit
glm-4.7-flashZhipu AIlatest~25.7 GBGood fit
north-mini-code-1.0Ollama Librarylatest~25.2 GBGood fit
qwen3-coderAlibaba Cloud30B~25.1 GBGood fit
qwen3.6Alibaba Cloud27B~23.6 GBGood fit
qwen3.8Alibaba Cloud27B~22.8 GBExcellent fit
muse-glimmerOllama Library30B~22.7 GBExcellent fit
mistral-small3.1Mistral AI24B~21.0 GBExcellent fit
devstral-small-2Mistral AI24B~20.6 GBExcellent fit
mistral-small3.2Mistral AI24B~20.6 GBExcellent fit
lfm2Ollama Library24B~19.6 GBExcellent fit
devstralMistral AI24B~19.5 GBExcellent fit
magistralMistral AI24B~19.5 GBExcellent fit
gpt-ossOpenAI20B~18.8 GBExcellent fit
gpt-oss-safeguardOpenAI20B~18.8 GBExcellent fit
mistral-smallMistral AI22B~18.2 GBExcellent fit
codestralMistral AI22B~18.2 GBExcellent fit
solar-proUpstage22B~18.1 GBExcellent fit
phi4-reasoningMicrosoft14B~15.2 GBExcellent fit
deepseek-coder-v2DeepSeek AI16B~14.2 GBExcellent fit

Showing the 25 largest of 194 models that run well.

How these results were calculated

Each model is evaluated at 4K context against 32GB of VRAM, counting model weights, KV cache, compute buffers and runtime overhead, minus a reserve for the display and operating system. Download sizes come from the official Ollama registry.

These figures assume 32GB of system RAM and working GPU drivers. Your own machine may differ — RAM, free disk space and whether an acceleration backend is actually installed all change the answer, which is what the PC scan measures directly.

GeForce RTX 5090 local AI — frequently asked questions

How many AI models can a GeForce RTX 5090 run?
Out of 228 models in the Ollama library, 194 run fully on a GeForce RTX 5090's 32GB of VRAM at 4K context, and a further 10 run with part of the model offloaded to system RAM, more slowly.
What is the largest AI model a GeForce RTX 5090 can run?
The heaviest build that fits entirely in VRAM is laguna-xs-2.1:latest, needing about 27.4 GB of memory from a 18.9 GB download.
Is 32GB of VRAM enough for local AI?
32GB is enough for 194 of the 228 models tested here, which covers most general-purpose and coding assistants. Very large models still need either partial CPU offload or a card with more memory.

Compare other graphics cards

GeForce RTX 5080GeForce RTX 5070 TiGeForce RTX 5070GeForce RTX 5060 Ti 16GBGeForce RTX 5060GeForce RTX 4090GeForce RTX 4080GeForce RTX 4070 TiGeForce RTX 4070GeForce RTX 4060 Ti 16GBGeForce RTX 4060 Ti 8GBGeForce RTX 4060GeForce RTX 3090GeForce RTX 3080GeForce RTX 3070GeForce RTX 3060 12GBGeForce RTX 3060 TiGeForce RTX 3050 8GB