Run a private ChatGPT-style UI for local models with Open WebUI

Description and use

Open WebUI provides a self-hosted ChatGPT-style UI for local models. Install with pip or Docker, then run the server or container. Pull a model with ollama pull llama3, open http://localhost:3000, and select the model. Keep the open-webui volume for persistent chat history and do not expose port 3000 publicly without a reverse proxy.

Problem

A user wants a polished, self-hosted chat interface that talks to local models without sending prompts or data to a hosted service.

Outcome

A private ChatGPT-style interface is running locally, talking to a local Ollama model, with chat history stored in a named volume.

Workflow

• Install Open WebUI via pip or Docker.
• Run the server or start the container.
• Pull a model with ollama pull llama3.
• Open http://localhost:3000 and create the first account as admin.
• Select the pulled model from the model dropdown in a new chat.

Evidence

You can get a ChatGPT-style interface that runs on your own machine and talks to local models.

Source

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

Source

ollama pull llama3

Source

Do not expose port 3000 to the internet without a reverse proxy and strong passwords.

Source