Privacy-First AI Interface
A clean, privacy-first interface for local AI models via Ollama, with flexible routing to external AI services.
100% free and open source. No telemetry, no tracking. Enhanced with Persona Development Framework and custom backgrounds.
Comprehensive AI Interface Features
Privacy by Default
Complete offline operation with local processing. Your conversations stay on your hardware unless you choose external connections.
207+ AI Models
Connect to 6 major AI providers: OpenAI (67 models), Anthropic (13), Groq (14), Gemini (45), Mistral (48), GitHub (20).
Interactive Artifacts
Render HTML, SVG, React components, and code directly in chat. Full-screen mode with copy/download capabilities.
Document Chat (RAG)
Upload PDF, TXT, DOCX, and Markdown files. Chat with your documents using semantic search and vector embeddings.
Enhanced User Management
Advanced authentication with role-based access control, optional password updates, and secure session management with JWT tokens.
Keyboard Shortcuts
VS Code-inspired shortcuts for power users. ⌘B (sidebar), ⌘D (theme), ⌘, (settings), ? (help).
Persona Development Framework
Create, manage, and apply custom AI personalities in chat sessions. Full persona creation with unique behaviors, visual identities, and specialized AI assistants.
Advanced Model Management
Enhanced model pulling with streaming progress, cancellation support, and optimized performance with SQLite database storage.
Open Source
Apache 2.0 licensed with transparent development. Inspect, modify, and distribute freely.
Get Started in Minutes
Comprehensive AI Platform
Interactive Artifacts
Create and interact with dynamic content directly within chat conversations:
- HTML Pages with JavaScript functionality
- SVG Graphics with animations
- Code snippets with syntax highlighting
- JSON data with validation
- Full-screen mode and download capabilities
Document Chat (RAG)
Upload documents and have intelligent conversations with your files:
- PDF, TXT, DOCX, Markdown support
- Vector embeddings for semantic search
- Smart chunking with context overlap
- Local processing with Ollama embeddings
- Persistent storage and management
Persona Development Framework
Create, manage, and apply custom AI personalities in chat sessions:
- Complete persona creation and editing system
- Persona selection in chat sessions
- JSON export/import for persona sharing
- Background application and personality traits
- Specialized AI assistant development
Custom Background Images
Personalize your interface with custom background images:
- Upload custom background images
- Real-time preview and blur effects
- Opacity control for text readability
- Toggle functionality and easy removal
- Persistent settings across sessions
Plugin System
Connect to external AI services while maintaining local fallback:
- OpenAI (67 models including o3, GPT-4o)
- Anthropic (13 models including Claude 4)
- Groq (14 models including Llama 4)
- Google Gemini (45 models)
- Mistral (48 models) & GitHub (20 models)
Complete Ollama Integration
Full integration with all Ollama API endpoints:
- Chat completion with streaming responses
- Multimodal support for vision models
- Model management (pull, delete, create)
- Embeddings generation for RAG
- Tool calling and structured outputs
Building AI with Integrity
Libre WebUI was created to provide a clean, reliable interface for AI interactions. We focus on user control, privacy by default, and practical functionality.
🛡️ Our Principles
- Open-source development with transparent governance
- Privacy by default, external connections by choice
- Inclusive community welcoming all contributors
- Practical functionality over flashy features
Our Commitment
Libre WebUI welcomes contributors from all backgrounds. We believe good software comes from diverse perspectives and collaborative development.
🌟 What We Value
Open collaboration from contributors of all backgrounds
Building tools that work well for real use cases
You decide how and where your AI interactions happen
Apache 2.0 licensed with transparent development
🎯 Our Focus
Clean, functional interfaces. We prioritize usability and reliability over complex features.
Privacy by default. Local processing with optional external connections when you need them.
Community collaboration. Development that welcomes diverse perspectives and contributions.
Good software serves its users, not the other way around.
Get Started with Privacy-First AI
Join users who value privacy and control in their AI interactions. Self-hosted, open source, and flexible.