Open WebUI

Open WebUI (formerly Ollama WebUI) is the world’s leading open-source, self-hosted AI chat interface with over 150,000 GitHub stars. Designed to run completely offline or connect with private cloud clusters, Open WebUI provides a ChatGPT-grade user experience with seamless support for Ollama, OpenAI-compatible APIs, local Retrieval-Augmented Generation (RAG), Python code execution, web browsing, and multi-user role management.

Core Enterprise & Self-Hosting Features

• Universal Model Provider Gateway: Native direct integration with Ollama (Llama 3.3, DeepSeek-R1, Mistral, Qwen) and OpenAI-compatible API endpoints (OpenAI, Anthropic Claude, Google Gemini, Groq, vLLM).
• Local RAG & Document Intelligence: Upload PDFs, CSVs, markdown docs, and text files. Open WebUI handles text chunking, embedding generation, and hybrid vector search directly in the browser.
• Web Search Integration: Live web groundings powered by SearXNG, Google PSE, Tavily, Brave Search, and DuckDuckGo for factual, up-to-date responses.
• Interactive Code Execution & Artifacts: Safely execute Python scripts in client-side or sandboxed environments, view live HTML/JS previews, and render interactive charts.
• Granular Role-Based Access Control (RBAC): Multi-user support with admin dashboards, model permission scoping, rate-limiting, and SSO authentication (OAuth, SAML).
• Pipelines & Custom Filter Functions: Extensible plugin architecture enabling custom pre/post-processing pipelines, rate limiting, and agent tool execution.

Quick Deployment with Docker

Deploy Open WebUI locally with full GPU acceleration in a single command:

# Run with default local Ollama support
docker run -d -p 3000:8080 --add-host=host.docker.internal:host-gateway -v open-webui:/app/backend/data --name open-webui --restart always ghcr.io/open-webui/open-webui:main

Frequently Asked Questions