Open WebUI is a self-hosted AI platform written in Python that provides a user-friendly web interface for interacting with large language models. The project supports multiple LLM runners including Ollama and OpenAI-compatible APIs, with a built-in inference engine for Retrieval Augmented Generation. It is designed to operate entirely offline while remaining extensible and feature-rich.
Open WebUI offers extensive integration capabilities, allowing users to connect any OpenAI-compatible API alongside local Ollama models and point to services like LMStudio, GroqCloud, Mistral, OpenRouter, and vLLM. The platform includes granular role-based access control with user groups and permissions, plugin support through Filters, Actions, Pipes, Tools, and Skills, and the ability to connect external services via MCP, MCPO, and OpenAPI tool servers. Users can wrap base models with custom instructions and tools to build specialized agents with dynamic variables and per-user or per-group access control.
The platform provides collaborative features including real-time shared channels where teams and AI models work together, persistent memory that carries context across conversations, and live workflow visualization for watching AI build checklists in real time. It includes a built-in calendar system with month, week, and day views, recurring events, color coding, and attendees, with models able to manage schedules conversationally through native function calling. Automation capabilities allow scheduling prompts on recurring schedules with runs surfaced on the calendar.
Open WebUI supports comprehensive document handling through local RAG integration backed by nine vector databases and multiple content-extraction engines including Tika, Docling, Document Intelligence, Mistral OCR, and PaddleOCR-vl. Web search functionality integrates dozens of providers including SearXNG, Google PSE, Brave Search, Kagi, and others. Image generation and editing capabilities support OpenAI DALL-E, Gemini, ComfyUI, and AUTOMATIC1111.
The platform supports multiple deployment options including pip, uv, Docker, and Kubernetes with specialized tagged images for Ollama and CUDA. It offers flexible database choices between SQLite with optional encryption or PostgreSQL, and file storage options including local storage, S3, Google Cloud Storage, and Azure Blob Storage. Enterprise features include LDAP and Active Directory integration, SSO via trusted headers and OAuth providers, SCIM 2.0 automated provisioning, and native Google Drive and OneDrive integration. The architecture supports horizontal scalability through Redis-backed session management and WebSocket support for multi-worker deployments. Open WebUI includes OpenTelemetry support for production observability and maintains multilingual support through i18n.
The project is part of a broader ecosystem that includes Open Terminal for code execution, Terminals for enterprise container isolation, cptr for mobile-first computing agents, oikb for knowledge base synchronization from 45+ sources, and a native desktop application for macOS, Windows, and Linux with system-wide Spotlight integration.
