Skip to content

Open WebUI Service

Open WebUI is a feature-rich, self-hosted web interface that provides a ChatGPT-like experience for interacting with various large language models. It offers an extensible platform supporting multiple LLM providers, local models, and advanced chat features.

Architecture Overview

graph TB
    A[Web Browser] -->|HTTPS/HTTP| B[Open WebUI Frontend]
    B -->|API Calls| C[Open WebUI Backend]
    C -->|Proxy Requests| D[LiteLLM Proxy]
    C -->|Direct Connection| E[Ollama]
    D --> F[OpenAI API]
    D --> G[Anthropic API]
    D --> H[Other Providers]
    C -->|Store Data| I[SQLite/PostgreSQL]
    C -->|File Storage| J[Local Storage]

    %% Styling
    classDef browserClass fill:#e8eaf6,stroke:#283593,stroke-width:2px,color:#000
    classDef frontendClass fill:#e1f5fe,stroke:#0277bd,stroke-width:2px,color:#000
    classDef backendClass fill:#e8f5e8,stroke:#2e7d32,stroke-width:2px,color:#000
    classDef proxyClass fill:#fff3e0,stroke:#ef6c00,stroke-width:2px,color:#000
    classDef localClass fill:#f3e5f5,stroke:#7b1fa2,stroke-width:2px,color:#000
    classDef apiClass fill:#ffebee,stroke:#c62828,stroke-width:2px,color:#000
    classDef storageClass fill:#fafafa,stroke:#424242,stroke-width:2px,color:#000

    class A browserClass
    class B frontendClass
    class C backendClass
    class D proxyClass
    class E localClass
    class F,G,H apiClass
    class I,J storageClass

Key Features

Multi-Provider Support

  • Seamless integration with multiple LLM providers
  • Model switching without service restart
  • Provider failover and load balancing
  • Custom API endpoint configuration

Advanced Chat Interface

  • ChatGPT-like conversational experience
  • Markdown rendering with syntax highlighting
  • LaTeX math equation support
  • Code execution capabilities
  • Image generation integration

File Management

  • Document upload and processing
  • PDF, DOCX, and text file support
  • Image upload for vision models
  • Conversation export (JSON, Markdown)

User Management

  • Multi-user support with authentication
  • Role-based access control
  • User registration and profile management
  • API key management

Customization Options

  • Custom model parameters (temperature, max tokens)
  • System prompts and templates
  • Custom CSS and theming
  • Plugin system for extensions

Configuration Schema

Environment Variables

# API Configuration
OLLAMA_BASE_URL=http://host.docker.internal:11434
OPENAI_API_BASE_URL=http://litellm:4000/v1
OPENAI_API_KEY=******************

# Security Configuration
WEBUI_SECRET_KEY=${WEBUI_SECRET_KEY}
WEBUI_JWT_SECRET_KEY=${WEBUI_JWT_SECRET_KEY}

# User Management
DEFAULT_USER_ROLE=admin
ENABLE_SIGNUP=true
ENABLE_LOGIN_FORM=true

# Model Configuration
DEFAULT_MODELS=gpt-3.5-turbo,gpt-4,claude-3-sonnet

# Interface Settings
WEBUI_NAME=AI Dev Local

Model Configuration Example

{
  "models": [
    {
      "id": "gpt-4",
      "name": "GPT-4",
      "provider": "openai",
      "parameters": {
        "temperature": 0.7,
        "max_tokens": 4096,
        "top_p": 1.0
      }
    },
    {
      "id": "claude-3-sonnet",
      "name": "Claude 3 Sonnet",
      "provider": "anthropic",
      "parameters": {
        "temperature": 0.5,
        "max_tokens": 2048
      }
    }
  ]
}

User Profile Schema

{
  "id": "user-123",
  "name": "John Doe",
  "email": "john@example.com",
  "role": "user",
  "preferences": {
    "theme": "dark",
    "language": "en",
    "default_model": "gpt-4",
    "show_timestamps": true
  },
  "api_keys": {
    "openai": "sk-...",
    "anthropic": "sk-ant-..."
  }
}

Default Configuration

  • Default Role: admin
  • Signup Enabled: Yes
  • Login Form: Enabled
  • Default Models: gpt-3.5-turbo, gpt-4

Access

Open WebUI is accessible at:

http://localhost:8081/

Supported Features

Model Providers

  • OpenAI: GPT-3.5, GPT-4, GPT-4 Vision
  • Anthropic: Claude 3 (Haiku, Sonnet, Opus)
  • Google: Gemini Pro, Gemini Pro Vision
  • Local Models: Via Ollama integration
  • Custom APIs: Any OpenAI-compatible endpoint

File Types

  • Documents: PDF, DOCX, TXT, MD
  • Images: PNG, JPG, GIF, WebP (for vision models)
  • Data: CSV, JSON, XML
  • Code: Various programming languages

Export Formats

  • Conversations: JSON, Markdown, PDF
  • Data: CSV for analytics
  • Images: Generated images and diagrams

Advanced Features

Function Calling

  • Built-in web search capabilities
  • Calculator and math functions
  • Weather and news APIs
  • Custom function definitions

Memory System

  • Conversation history storage
  • Cross-session memory retention
  • Searchable chat history
  • Context-aware responses

Plugin System

  • Community-developed plugins
  • Custom integration capabilities
  • API extensions
  • Third-party service connectors

Online Resources

Use Cases

  • Personal AI Assistant: Private ChatGPT-like interface
  • Team Collaboration: Shared AI workspace for teams
  • Development Tool: Code generation and debugging assistant
  • Research Platform: Academic and professional research aid
  • Customer Support: AI-powered support interface
  • Content Creation: Writing and creative assistance

Security Features

  • Authentication: JWT-based user authentication
  • Authorization: Role-based access control
  • Data Privacy: Local data storage options
  • API Security: Secure API key management
  • CORS Protection: Configurable CORS policies

Open WebUI is ideal for individuals and organizations seeking a powerful, self-hosted alternative to commercial AI chat interfaces, with full control over data and customization options.