Open WebUI: Deploy a ChatGPT-Like Interface for Your Local Models

Updated
7 min read
Open WebUI: Deploy a ChatGPT-Like Interface for Your Local Models
TL;DR

Open WebUI gives you a full-featured, self-hosted chat interface with RAG, multi-user support, and image generation - running entirely on your infrastructure.

What Is Open WebUI

Open WebUI is a self-hosted web interface for local LLMs, designed as a feature-complete alternative to ChatGPT. It connects to Ollama (or any OpenAI-compatible endpoint) and adds a polished UI, document-based RAG, image generation, multi-user management, and a mobile-friendly PWA - all containerised and easy to deploy.

One-Command Docker Install With Ollama

If Ollama is already running on your machine:

bash
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

Open http://localhost:3000 and sign in with the admin account created on first launch. Open WebUI auto-discovers models from Ollama at host.docker.internal:11434.

Docker Compose With Bundled Ollama

bash
version: "3.8"
services:
  ollama:
    image: ollama/ollama:latest
    volumes:
      - ollama:/root/.ollama
    ports:
      - "11434:11434"

  open-webui:
    image: ghcr.io/open-webui/open-webui:main
    depends_on:
      - ollama
    ports:
      - "3000:8080"
    environment:
      - OLLAMA_BASE_URL=http://ollama:11434
    volumes:
      - open-webui:/app/backend/data
    restart: always

volumes:
  ollama:
  open-webui:

Model Management UI

From the admin panel you can pull, delete, and switch models without touching the terminal. Open WebUI shows model size, VRAM usage, and pull progress in real time.

RAG Pipeline

Upload PDFs, Word docs, or web URLs in any conversation. Open WebUI chunks the content, embeds it with a local embedding model (default: nomic-embed-text via Ollama), stores vectors in its built-in ChromaDB, and injects relevant chunks into the system prompt automatically. The docs cover configuring external vector databases and adjusting chunk size and overlap.

Image Generation via AUTOMATIC1111

Connect an AUTOMATIC1111 Stable Diffusion WebUI instance in Settings → Images to enable /imagine prompts inside chat conversations. The generated image appears inline in the message thread.

Multi-User With Roles

Open WebUI has a built-in user database with Admin, User, and Pending roles. Admins can control which models each user can access, set per-user API key auth, and view usage statistics. This makes it suitable for small-team internal deployments.

PWA on Mobile

Open WebUI ships as a Progressive Web App. Add it to your phone's home screen from Safari or Chrome for a native-app-like experience with persistent chat history and offline model caching.

API Key Authentication

For programmatic access, generate API keys from your profile settings. The API mirrors the OpenAI format - useful for connecting external tools to your self-hosted model stack without exposing Ollama directly.

#open-webui#ollama#self-hosted#rag#docker

One AI engineering post, weekly

LLM benchmarks, prompt techniques, and token-cost breakdowns, delivered to your inbox.

What should we test or explain next?

Comments

Have a question or something to add? Sign in to join the discussion.

Mahmudul Haque Qudrati — CEO & ML Engineer at Pristren

Mahmudul Haque Qudrati

CEO & ML Engineer

Software architect, full-stack engineer, and machine learning practitioner. Founder & CEO of Pristren, CTO at Venture Kick-backed PIN Sport Application GmbH (Basel, Switzerland), Software Development Lead at DailyFrame GmbH (building Dimetra patient transport software), and creator of Zlyqor. Focused on custom workflow automation, scalable web and mobile applications, and production AI/ML systems.