Skip to content
AI Stack Nest
Windows2 min read · updated 17 Nov 2026

Install Ollama on Windows and run your first model

Step-by-step: check requirements, install Ollama on Windows, run a first model, move model storage and fix common problems.

Before you start

Ollama runs AI models on your own computer, so nothing you type is sent to a cloud service by the model itself. According to the official Ollama documentation, Windows needs Windows 10 22H2 or newer. If you have an NVIDIA graphics card, the documented minimum driver is version 551.61 or newer. AMD cards are supported through ROCm v7 or Vulkan.

You also need free disk space. The documentation asks for at least 4 GB for the Ollama program and then more for each model. A small 7B or 8B model is roughly 4 to 5 GB (Llama 3.1 8B is about 4.9 GB, Qwen2.5 7B about 4.7 GB), so plan on 10 GB free to start comfortably.

Step 1: Download and install

  1. Open the official download page at ollama.com/download and choose Windows.
  2. Run the file called OllamaSetup.exe.
  3. The installer works inside your user account and does not need administrator rights.

If you want Ollama installed on another drive, start the installer from a terminal with the /DIR option, for example OllamaSetup.exe /DIR="d:\some\location".

Step 2: Run a model

Open PowerShell or Command Prompt and run:

ollama run llama3.1

The first run downloads the model, which can take a while. When the >>> prompt appears, type a question. Type /bye to leave the chat.

Step 3: Check that it is working

Ollama listens on http://localhost:11434. Run ollama list to see the models you have installed and ollama ps to see which are loaded and whether they run on the GPU or the CPU.

Where models are stored

Models are saved in %HOMEPATH%\.ollama by default. To move them to a bigger drive, set the OLLAMA_MODELS environment variable to the new folder and restart Ollama.

Troubleshooting

  • The model is slow. Run ollama ps. If it says CPU, your GPU is not being used. Update your graphics driver and check it meets the minimum version above.
  • Not enough disk space. Delete models you no longer need with ollama rm MODEL.
  • Behind a company proxy. The Ollama FAQ says to set HTTPS_PROXY; avoid HTTP_PROXY because it can break model downloads.

Sources

Facts on this page come from the official Ollama documentation for Windows and its FAQ, read on 2026-11-17.

Models used here

More setup guides

Not sure your computer can run a model? Use the will-it-run checker.