Private AI Assistant
Ollama & Open-WebUI

Run large language models like Llama 3, DeepSeek, and Mistral entirely on your own server hardware with zero subscription fees or data leaks.

Why Self-Host Your AI?

Commercial AI services are convenient, but sending proprietary documents or private codebase logic to third-party APIs carries privacy risks. By combining Ollama (the backend inference engine) with Open-WebUI (a web frontend interface), you get your own ChatGPT alternative running 100% on your infrastructure.

1. Installing Ollama Inference Engine

Install Ollama on your Linux host using the official installation script:

curl -fsSL https://ollama.com/install.sh | sh

Download and test your first open-source language model (e.g. Llama 3 or DeepSeek-Coder):

ollama run llama3

2. Deploying Open-WebUI via Docker

Deploy Open-WebUI connected directly to your host's Ollama instance:

docker run -d -p 3000:8080 --gpus=all \
  --add-host=host.docker.internal:host-gateway \
  -v open-webui:/app/backend/data \
  --name open-webui \
  --restart unless-stopped \
  ghcr.io/open-webui/open-webui:main

Open http://your-server-ip:3000 in your browser to create your initial administrator account.

Why Self-Host Your AI?

1. Installing Ollama Inference Engine

2. Deploying Open-WebUI via Docker

Related AI & Docker Guides

Stable Diffusion AI Generator

Docker & Portainer Setup