Run Open WebUI with Ollama and Docker Desktop on Windows

Do not worry about your difficulties in mathematics. I can assure you mine are still greater, Albert Einstein

Ollama is a lightweight, privacy-focused platform that lets you run large language models (LLMs) locally on your own machine —no cloud dependency or costly monthly subscriptions required. It’s designed to make working with models like Llama 3, DeepSeek, Gemma, and others as simple as running a command in your terminal.

Docker Desktop is an application that provides a seamless, integrated environment for building, sharing, and running containerized applications on your local machine running Windows. Combined with Ollama and Open WebUI, you get a full local-AI stack:

1. Ollama: CLI for pulling and running LLMs on your machine.
2. Docker Desktop: Container engine and GUI to manage your apps.
3. Open WebUI: A ChatGPT-style web interface to interact or chat with your local models.

You may also want to read our article Complete Windows AI Dev Setup: WSL 2, Docker Desktop, Python & Ollama, a step-by-step guide to build a modern AI development workstation on Windows.

Install Docker Desktop & Ollama

1. Install Docker Desktop. Download the latest Windows installer from docker.com, run the installer, follow the Installation wizard prompts, and reboot if prompted. Finally, open Docker Desktop and sign in to your Docker Hub account.

2. Install Ollama. Visit https://ollama.com and download the Windows installer. Run the .exe, follow the Installation wizard prompts, then reboot (if required). Verify with: ollama version.

Open WebUI

Open WebUI is the best open-source, user‑friendly web interface for running local AI models. It sits on top of Ollama. This comprehensive guide will walk you through installation, configuration, and getting your first AI model running locally.

Open WebUI provides a ChatGPT-like interface for interacting with local AI models through Ollama. It offers:

Privacy-focused: All conversations and data stay on your machine -what happens in Vegas stays in Vegas!
Model flexibility: Support for various Ollama-supported models.
User-friendly: A very intuitive, ChatGPT-style web interface similar to popular and commercial AI chatbots.
Customizable Themes & Multi-User Support, e.g., Settings, General, Choose theme for Docker Desktop: Dark.

You need: Docker Desktop installed and running, sufficient disk space, and some hardware requirements: 4+ cores, 8GB/16GB RAM, NVIDIA with CUDA, 16 GB+ RAM (GPU-accelerated), and 20+ GB free space for models and containers.

Installation Steps

Step 1: Pull the Docker Image based on your hardware. For NVIDIA GPU with CUDA support: docker pull ghcr.io/open-webui/open-webui:cuda. For CPU-only setups: docker pull ghcr.io/open-webui/open-webui:main.

Step 2: Run the Container. With NVIDIA GPU support: docker run -d -p 3000:8080 --gpus all -v open-webui:/app/backend/data --name open-webui ghcr.io/open-webui/open-webui:cuda. CPU-only setup: docker run -d -p 3000:8080 -v open-webui:/app/backend/data --name open-webui ghcr.io/open-webui/open-webui:main (CPU-only).

Step 3: Verify Installation. Open Docker.desktop, navigate to Containers, and confirm open-webui shows as Running. Click on the port link or visit `[http://] + localhost + [:3000].

Step 4. Initial Setup On first visit, you’ll be prompted to create an admin account (name, email, and password), then access the interface.

Adding Your First Model

1. Navigate to Models: In the WebUI, click on “Models” in the sidebar.

2. Add New Model: Click “Add Model” or the “+” button.

3. Enter Model Identifier: Type the model name, e.g., deepseek-r1:14b, codellama:7b, deepseek-r1:14b, etc.

4. Pull Model: Select “Pull [model-name] from Ollama.com”.

5. Wait for Download: Monitor the download progress (this may take several minutes or more) and wait for the download to complete.

   

Performance Optimization

• Use GPU acceleration when available: Ensure Docker Desktop has GPU enabled.
• Allocate sufficient RAM to Docker. In Docker settings, allocate 8 GB+ for WSL.
• Consider using smaller, quantized models (a cornerstone technique for deploying large language models and deep learning systems in resource-constrained environments.)
• Enable model caching for frequently used models. Keep frequently used models loaded.

How To Change Ollama Model Default Directory To Save Your Hard Drive Storage

By default, Ollama stores models in %USERPROFILE%.ollama\models. To move them:

1. To change the directory where Ollama stores downloaded models on Windows, you need to set the environment variable OLLAMA_MODELS. Search for environment variables in the Windows search bar and select “Edit the system environment variables”. In the System Properties window, click on the Environment Variables… button.
2. Create a New Environment Variable. In the Environment Variables window, look for the System variables section and click on the New…. button.

   You can set variables for your specific user (“User variables”) or for all users (“System variables”).

3. Enter OLLAMA_MODELS for the Variable name and type the path where you want to save the model files for the Variable value (e.g., D:\ollama\models), and click OK on all open windows to save the changes.
4. If Ollama is running, close it (right-click the Ollama icon in the system tray and select “Quit”) and then restart it.
5. Download a new model and check if it is saved in the new directory.

Install Portainer using Docker Desktop

Portainer is a GUI for Docker management—view containers, images, volumes, networks, and more.

1. Download and Install Docker Desktop. If you haven’t already, download and install Docker Desktop from the official Docker website.
2. Install Portainer Extension. Open Docker Desktop and navigate to the Extensions marketplace (usually accessible from the left-hand navigation panel) within the Docker Desktop, click on Browse, search for Portainer and click Install. Follow the prompts to complete the installation.
3. Access Portainer. Once the installation is complete, you can access Portainer through Docker Desktop. Look for the Portainer icon in the Extensions section and click to open it.

With this stack —Ollama, Docker Desktop, and Open WebUI— you have a fully local, private, and powerful AI playground on Windows.