Local AI on Windows

Written by Jakub Rusinowski · Last updated July 21, 2026

The widest hardware choice and the easiest NVIDIA setup, at the cost of the highest idle memory overhead of the three platforms.

Runtimes that work on Windows

RuntimeHow to installAccelerationWatch out for
OllamaOfficial Windows installer (.exe), or winget.
winget install Ollama.Ollama
CUDA (NVIDIA), ROCm (select AMD), Vulkan, CPUROCm support on Windows covers far fewer AMD cards than on Linux; check your specific model before relying on it.
LM StudioOfficial Windows installer, GUI-first with a built-in model browser.CUDA (NVIDIA), Vulkan, CPU
llama.cppPrebuilt release binaries, or build from source with CUDA enabled.
winget install llama.cpp
CUDA (NVIDIA), Vulkan, SYCL (Intel), CPUPrebuilt Windows binaries are split by backend — download the CUDA build, not the CPU one, or you will silently run on CPU.
vLLMRuns under WSL2, not natively on Windows.
wsl --install
CUDA (NVIDIA, via WSL2)WSL2 splits memory with Windows and adds a virtualisation layer; for single-user local inference Ollama or llama.cpp is the simpler choice.

What Windows is good at

Windows limitations

Hardware and models on Windows

Windows supports NVIDIA, AMD, Intel accelerators — 37 of the 55 in our hardware database, up to 128 GB.

MemoryHardware on this platformModels to run
8 GBNVIDIA GeForce RTX 3080 (10GB)
Intel Arc B570
NVIDIA GeForce RTX 5060 Ti 8GB
Qwen 3 8B
GLM-6 9B
GLM-4.7 9B
16 GBAMD Radeon RX 7900 XT
NVIDIA GeForce RTX 5080
NVIDIA GeForce RTX 5070 Ti
Mistral Small 3.1 24B
Qwen 3.5 14B
Qwen 3.5 14B
24 GBNVIDIA GeForce RTX 4090
NVIDIA GeForce RTX 3090
AMD Radeon RX 7900 XTX
Gemma 4 27B ⭐
Mistral Small 3.1 24B
Qwen 3.7 35B-A3B
48 GBNVIDIA RTX 6000 Ada Generation
NVIDIA L40S
Nemotron 70B Instruct
GLM-5.1 72B
Qwen 3.5 72B
96 GBAMD Ryzen AI Max+ 395
NVIDIA RTX PRO 6000 Blackwell
GPT-oss 120B
Nemotron 70B Instruct
Qwen 3.5 122B-A10B (MoE)

Machines that run Windows

MachineMemory for modelsForm factorUpgradeable
RTX 4090 Desktop (24 GB VRAM, 64 GB RAM)24 GB VRAMDesktopYes
RTX 5090 Desktop (32 GB VRAM, 64 GB RAM)32 GB VRAMDesktopYes
RTX 3090 Desktop (24 GB VRAM, 64 GB RAM)24 GB VRAMDesktopYes
RTX 5080 Desktop (16 GB VRAM, 32 GB RAM)16 GB VRAMDesktopYes
RTX 3060 12 GB Desktop (12 GB VRAM, 32 GB RAM)12 GB VRAMDesktopYes
RTX 5060 Ti 16 GB Desktop (16 GB VRAM, 32 GB RAM)16 GB VRAMDesktopYes
Framework Desktop (Ryzen AI Max+ 395, 128 GB)128 GB unifiedMini PCNo
RTX 4060 Laptop (8 GB VRAM, 16 GB RAM)8 GB VRAMLaptopYes

FAQ

What is the best way to run an LLM locally on Windows?

Ollama — Official Windows installer (.exe), or winget. It reaches CUDA (NVIDIA), ROCm (select AMD), Vulkan, CPU.

Which GPUs work for local AI on Windows?

NVIDIA, AMD, Intel hardware, up to 128 GB in our database.

What are the downsides of running local AI on Windows?

Windows itself reserves noticeably more VRAM and system RAM than Linux, so usable memory for a model is lower on identical hardware.

Machines Running This Platform

Hardware for This Platform

Other Platforms

Tools

← All platforms | Check your hardware