Open WebUI

This guide covers the installation, configuration, and usage of Open WebUI with Ollama, providing a user-friendly graphical interface for interacting with your local large language models.

What is Open WebUI?

Open WebUI (formerly Ollama WebUI) is an open-source web interface designed specifically for Ollama. It provides:

• A chat-like interface for interacting with models

• File upload and RAG capabilities

• Multi-modal support (text and images)

• Vision features for supported models

• Prompt templates and history

• User management

• API integrations

• Custom model configurations

Prerequisites

Before installing Open WebUI, ensure you have:

1. A working Ollama installation (follow the installation guide)

2. Docker and Docker Compose (recommended for easy setup)

3. 4GB+ RAM available (in addition to what Ollama requires)

4. At least one model installed in Ollama

5. Ollama running and accessible on port 11434

Installation Methods

Method 1: Docker (Recommended)

The easiest way to install Open WebUI is using Docker:

# Pull the latest Open WebUI image
docker pull ghcr.io/open-webui/open-webui:latest

# Run Open WebUI container connecting to local Ollama
docker run -d --name open-webui \
  -p 3000:8080 \
  -e OLLAMA_API_BASE_URL=http://host.docker.internal:11434/api \
  -v open-webui-data:/app/backend/data \
  ghcr.io/open-webui/open-webui:latest

Access the interface at http://localhost:3000

Method 2: Docker Compose

Create a docker-compose.yml file containing both Ollama and Open WebUI:

version: '3'

services:
  ollama:
    image: ollama/ollama:latest
    container_name: ollama
    volumes:
      - ollama-data:/root/.ollama
    ports:
      - "11434:11434"
    restart: unless-stopped

  open-webui:
    image: ghcr.io/open-webui/open-webui:latest
    container_name: open-webui
    volumes:
      - open-webui-data:/app/backend/data
    ports:
      - "3000:8080"
    environment:
      - OLLAMA_API_BASE_URL=http://ollama:11434/api
    depends_on:
      - ollama
    restart: unless-stopped

volumes:
  ollama-data:
  open-webui-data:

Deploy with:

docker compose up -d

Access the interface at http://localhost:3000

Method 3: Manual Installation

For users who prefer not to use Docker:

# Clone the repository
git clone https://github.com/open-webui/open-webui.git
cd open-webui

# Install backend dependencies
cd backend
python -m venv venv
source venv/bin/activate
pip install -r requirements.txt

# Configure environment
cp .env.example .env
# Edit .env to set OLLAMA_API_BASE_URL=http://localhost:11434/api

# Start the backend
uvicorn main:app --host 0.0.0.0 --port 8080

# In another terminal, install and start the frontend
cd ../frontend
npm install
npm run dev

Access the interface at the URL provided by the frontend development server.

Configuration Options

Environment Variables

Open WebUI can be configured with various environment variables:

Variable	Description	Default
OLLAMA_API_BASE_URL	URL of Ollama API	http://localhost:11434/api
PORT	Port for the web interface	8080
HOST	Host binding for the interface	0.0.0.0
DATA_DIR	Directory for storing data	/app/backend/data
ENABLE_SIGNUP	Allow new users to register	true
ENABLE_AUTH	Enable authentication	false
LOG_LEVEL	Logging detail level	error

For Docker deployments, pass these as environment variables:

docker run -d --name open-webui \
  -p 3000:8080 \
  -e OLLAMA_API_BASE_URL=http://host.docker.internal:11434/api \
  -e ENABLE_AUTH=true \
  -e LOG_LEVEL=info \
  -v open-webui-data:/app/backend/data \
  ghcr.io/open-webui/open-webui:latest

Authentication

To enable multi-user authentication:

1. Set ENABLE_AUTH=true in your environment variables

2. Set ENABLE_SIGNUP=true to allow new user registration (can be disabled later)

3. Open the web interface and create your first admin user

4. Set ENABLE_SIGNUP=false to prevent further registrations if desired

Persistent Storage

For persistent storage of conversations, settings, and users:

docker run -d --name open-webui \
  -p 3000:8080 \
  -e OLLAMA_API_BASE_URL=http://host.docker.internal:11434/api \
  -v ./open-webui-data:/app/backend/data \
  ghcr.io/open-webui/open-webui:latest

Using Open WebUI

Initial Setup

1. Open your browser and navigate to http://localhost:3000 (or your configured port)

2. If authentication is enabled, create an account or log in

3. In the sidebar, you should see available models from your Ollama instance

4. If no models appear, check that Ollama is running and accessible

Basic Chat Interface

The Open WebUI interface includes:

• Left sidebar: Models, conversations, and settings

• Main chat area: Messages between you and the model

• Input area: Text field for sending prompts to the model

• Model settings: Configuration panel for adjusting model parameters

Advanced Features

Custom Model Parameters

To customize model parameters for a specific conversation:

1. Click on the model name in the top bar of the chat

2. Adjust parameters:

   • Temperature: Controls randomness (0.0-2.0)

   • Top P: Nucleus sampling threshold (0.0-1.0)

   • Maximum length: Limits response length

   • Context window: Sets available context tokens

3. Save settings to apply them to the current conversation

RAG (Retrieval-Augmented Generation)

Enable RAG capabilities for improved responses with external knowledge:

1. In the sidebar, navigate to "RAG" section

2. Click "Upload files" to add documents (PDFs, text files, etc.)

3. Create a new collection and add your documents

4. When chatting, toggle the RAG feature to use your document collection

5. Ask questions related to your documents to see context-aware responses

Chat Templates

Create templates for common prompts:

1. Navigate to "Templates" in the sidebar

2. Click "New Template"

3. Define your template with placeholder variables

4. Save and use these templates in conversations

Vision Support

For models that support image input (like LLaVA):

1. Ensure you have a multimodal model like llava installed

2. In the chat interface, click the upload button (📎)

3. Select an image to upload

4. Ask questions about the image

DevOps Team Setup

For teams using Ollama in DevOps workflows:

Collaborative Setup

version: '3'

services:
  ollama:
    image: ollama/ollama:latest
    container_name: ollama
    volumes:
      - ollama-data:/root/.ollama
      - ./modelfiles:/modelfiles
    ports:
      - "11434:11434"
    restart: unless-stopped

  open-webui:
    image: ghcr.io/open-webui/open-webui:latest
    container_name: open-webui
    volumes:
      - open-webui-data:/app/backend/data
    ports:
      - "3000:8080"
    environment:
      - OLLAMA_API_BASE_URL=http://ollama:11434/api
      - ENABLE_AUTH=true
      - ENABLE_SIGNUP=false
    depends_on:
      - ollama
    restart: unless-stopped

  caddy:
    image: caddy:latest
    container_name: caddy
    restart: unless-stopped
    ports:
      - "80:80"
      - "443:443"
    volumes:
      - ./Caddyfile:/etc/caddy/Caddyfile
      - caddy-data:/data
      - caddy-config:/config
    depends_on:
      - open-webui

volumes:
  ollama-data:
  open-webui-data:
  caddy-data:
  caddy-config:

With a Caddyfile:

ollama.example.com {
    reverse_proxy open-webui:8080
    tls internal
}

Custom DevOps Modelfile

Create a special Modelfile for your team:

# Modelfile - Save as ./modelfiles/DevOpsAssistant
FROM codellama:latest

# Configure parameters
PARAMETER temperature 0.2
PARAMETER top_p 0.9

# System prompt
SYSTEM You are an expert DevOps assistant specialized in infrastructure as code, 
CI/CD pipelines, cloud platforms, and Kubernetes. You provide practical solutions 
for DevOps challenges with secure, modern best practices.
You focus on AWS, Azure, and GCP platforms, helping with Terraform, CloudFormation,
Bicep, Docker, Kubernetes, GitHub Actions, and Azure DevOps.

Build the model:

docker exec -it ollama ollama create devops-assistant -f /modelfiles/DevOpsAssistant

Security Considerations

When deploying Open WebUI:

1. Authentication: Always enable authentication in production

2. Network access: Limit access using a reverse proxy with TLS

3. User management: Control who has access to the interface

4. Document handling: Be aware that uploaded documents are stored in the data volume

5. API security: Protect the Ollama API endpoint from unauthorized access

Securing with Nginx

Example nginx.conf for securing Open WebUI:

server {
    listen 80;
    server_name ollama.example.com;
    return 301 https://$host$request_uri;
}

server {
    listen 443 ssl;
    server_name ollama.example.com;

    ssl_certificate /etc/nginx/certs/cert.pem;
    ssl_certificate_key /etc/nginx/certs/key.pem;

    # Basic authentication (optional additional layer)
    auth_basic "Restricted Access";
    auth_basic_user_file /etc/nginx/.htpasswd;

    location / {
        proxy_pass http://open-webui:8080;
        proxy_set_header Host $host;
        proxy_set_header X-Real-IP $remote_addr;
        proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
        proxy_set_header X-Forwarded-Proto $scheme;
    }
}

Troubleshooting

Issue	Solution
"Cannot connect to Ollama" error	Verify Ollama is running: curl http://localhost:11434/api/tags
Models not appearing	Ensure OLLAMA_API_BASE_URL is correctly set
File uploads failing	Check that the data directory is writable
Authentication issues	Clear browser cache or check user database
Slow performance	Adjust model parameters or upgrade hardware

Extending Open WebUI

API Access

Open WebUI provides its own API that can be accessed at:

http://localhost:3000/api/docs

This Swagger UI allows for programmatic interaction with the interface.

Custom Plugins

You can extend Open WebUI with custom plugins:

1. Clone the repository

2. Create a new directory in backend/plugins/

3. Implement the plugin interface

4. Add your plugin to the configuration

Next Steps

After setting up Open WebUI:

1. Customize models for specific use cases

2. Optimize GPU acceleration for better performance

3. Implement DevOps workflows using the web interface

4. Create team-specific templates and RAG collections

Additional Resources

• Open WebUI GitHub Repository

• Ollama Documentation

• Docker Hub