Open WebUI (formerly Ollama WebUI) is a self-hosted, feature-rich web interface for local AI models. It transforms Ollama’s command-line models into a polished chat experience comparable to ChatGPT or Claude.ai — but entirely on your own hardware, with no data leaving your machine. This guide covers installation, configuration, and the most useful features.

What Open WebUI Provides

ChatGPT-style UI — persistent conversations, model switching, markdown rendering
Multi-model support — chat with any Ollama model or OpenAI API
Multi-user — create accounts for family or team members, each with separate conversation history
RAG (Retrieval-Augmented Generation) — upload documents and chat with them
Web search integration — connect to SearXNG or other search APIs
Image generation — connect to AUTOMATIC1111 or ComfyUI
Voice input/output — speech-to-text and text-to-speech
Model management — download and manage Ollama models from the UI

Prerequisites

Docker and Docker Compose installed
Ollama running on the same machine or accessible over the network

If you haven’t set up Ollama:

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

# Pull a model
ollama pull llama3.1:8b
ollama pull mistral:7b

Installing Open WebUI with Docker

The easiest installation is via Docker Compose. Create a docker-compose.yml:

services:
  open-webui:
    image: ghcr.io/open-webui/open-webui:main
    container_name: open-webui
    volumes:
      - open-webui:/app/backend/data
    ports:
      - "3000:8080"
    environment:
      - OLLAMA_BASE_URL=http://host.docker.internal:11434
    extra_hosts:
      - host.docker.internal:host-gateway
    restart: unless-stopped

volumes:
  open-webui:

Start it:

docker compose up -d

Open http://localhost:3000 in your browser. The first user to sign up becomes the admin.

Linux (without Docker)

pip install open-webui
open-webui serve

Access at http://localhost:8080.

Initial Configuration

Click Sign Up and create your admin account
Select a model from the dropdown (your Ollama models appear automatically)
Start chatting

Admin Settings

Go to Admin Panel (user menu top-right) → Settings:

General: Set the default model, enable/disable signups
Models: Set which models are visible to users
Connections: Add additional model backends (OpenAI API, LM Studio, etc.)
Documents: Configure the RAG pipeline

Connecting Multiple AI Backends

Open WebUI can connect to multiple backends simultaneously:

Adding OpenAI API

Settings → Connections → OpenAI API:

URL: https://api.openai.com/v1
API Key: your OpenAI API key

Now GPT-4o and other OpenAI models appear in the model dropdown alongside your local Ollama models.

Adding LM Studio

LM Studio’s local server is OpenAI-compatible:

URL: http://localhost:1234/v1
API Key: not-needed

Adding a Remote Ollama Instance

If Ollama runs on a home server but you want to use Open WebUI from another machine:

URL: http://192.168.1.100:11434 (your server’s IP)

RAG: Chat With Your Documents

Open WebUI includes built-in RAG (Retrieval-Augmented Generation) for querying your own documents:

Click the + button in a new chat
Select Upload Files
Upload PDFs, Word docs, or text files
Ask questions: “Summarize this document” or “What does it say about X?”

The system chunks the document, embeds it with a local embedding model, and retrieves relevant sections to answer your questions. This is entirely local — your documents never leave your machine.

Setting Up a Knowledge Base

For persistent document collections:

Workspace → Knowledge → Create Knowledge
Name it (e.g., “Company Wiki”)
Add documents to the collection
In a chat, click # and select the knowledge base to activate it

Now the AI has access to your entire document collection as context.

Web Search Integration

Connect Open WebUI to a search engine for real-time information:

Admin Panel → Settings → Web Search:

Enable web search
Connect to SearXNG: http://localhost:8080 (if you have SearXNG running)
Or use Tavily API, Brave Search API, or DuckDuckGo

When enabled, users can activate web search per-chat with the globe icon.

Model Management

Pull new Ollama models directly from the Open WebUI interface:

Go to Admin Panel → Settings → Models
Click Download a model from Ollama.com
Type a model name: phi4:14b or qwen2.5:7b
Click Pull — downloads and installs

No command-line access needed.

Multi-User Setup

For a household or small team:

Enable user registration (or use invite links for private deployment)
Each user gets their own conversation history
Admins can set per-user model access
Share the URL (http://your-server-ip:3000) — works from any device on the LAN

Useful Settings for Daily Use

Model configuration per-chat: Click the gear icon in a chat to adjust temperature, max tokens, and system prompt without changing global defaults.

Custom system prompts per model: Set in Workspace → Models → Edit a model → System Prompt. This creates a pre-configured version of a model (e.g., “Coding Assistant”, “Writing Editor”) in the model dropdown.

Export conversations: Individual conversations can be exported as JSON or markdown via the chat menu.

Keyboard shortcuts: Ctrl+Enter sends the message. Shift+Enter adds a newline.

Open WebUI transforms a command-line Ollama installation into a full-featured AI platform that’s genuinely pleasant to use daily. For households running a home server, it’s an excellent way to give everyone access to local AI with proper accounts and history, all without internet connectivity or API costs.