Ollama on Windows: Requirements, Moving Models to Another Drive, and Where Files Live

By Web Asset House Studio · October 29, 2026

Note: Some future links on this site may be affiliate links. We never claim hands-on testing; guides are researched from the public sources listed in each article.

How we wrote this: A summary of Ollama's official Windows documentation (linked below), checked October 29, 2026. Driver and version requirements change, so check the docs before you install.

Ollama runs as a native Windows app. Once installed, it runs in the background and the ollama command works in Command Prompt, PowerShell or any terminal. This guide covers the Windows-specific parts. For a first model, see our beginner walkthrough.

Requirements

From Ollama's Windows page:

• Windows 10 22H2 or newer, Home or Pro.
• NVIDIA cards: driver version 551.61 or newer.
• AMD Radeon cards: a ROCm v7 / HIP7-capable driver for ROCm acceleration, or a Vulkan-capable Radeon driver for Vulkan acceleration. The docs note that some Radeon RX 6000-class systems may not expose ROCm v7 on current Windows drivers; Vulkan is on by default and is the recommended fallback.
• Disk space: at least 4 GB for the program itself, plus room for models, which the docs say "can be tens to hundreds of GB." To estimate which models fit your computer's memory, try the local AI calculator on our sister site.

If you are still on Windows 10, note that Microsoft's support timeline matters more than Ollama's here. See our guide to your options after Windows 10 end of support.

Installing without admin rights

The OllamaSetup.exe installer doesn't need Administrator access and installs into your home folder by default. To install somewhere else, the docs show a flag:

OllamaSetup.exe /DIR="d:\some\location"

Move models to a bigger drive

Models are the big files. By default they go in your home folder. To store them elsewhere, set the OLLAMA_MODELS environment variable for your account. The docs' steps:

1. Open Settings (Windows 11) or Control Panel (Windows 10) and search for environment variables.
2. Click Edit environment variables for your account.
3. Create or edit a variable named OLLAMA_MODELS and set it to the folder you want, for example D:\ollama-models.
4. Click OK/Apply.

If Ollama is already running, quit it from the system tray and start it again from the Start menu (or open a new terminal), so it picks up the change.

Where Ollama keeps its files

Press Win+R and paste one of these paths, as listed in the docs' troubleshooting section:

server.log holds the most recent server logs, so it's a good first stop when something goes wrong.

Calling the API from PowerShell

The local API runs at http://localhost:11434, the same as on Mac and Linux. The docs give this PowerShell example:

(Invoke-WebRequest -method POST -Body '{"model":"llama3.2", "prompt":"Why is the sky blue?", "stream": false}' -uri http://localhost:11434/api/generate ).Content | ConvertFrom-json

More on the API: Ollama API basics.

Uninstalling

The installer registers an uninstaller, so remove Ollama from Add or remove programs in Windows Settings. The docs warn that if you moved your models with OLLAMA_MODELS, the uninstaller won't delete them, so remove that folder yourself if you want the space back.

Running it as a service

For running Ollama as a background service or embedding it in another app, the docs describe a standalone ollama-windows-amd64.zip with just the CLI and NVIDIA libraries, plus separate packages for AMD (ROCm) and MLX (CUDA). You run ollama serve with a service tool of your choice. Most home and office users should stick with the regular installer, which also keeps Ollama updated.

Sources

• Ollama: Windows