Open WebUI

⭐ 128.0k MIT Python/TypeScript 0.6.0

A self-hosted AI office platform that integrates RAG knowledge bases with local document processing to enable intelligent work in a completely offline environment.

📋 Info

GitHub Stars⭐ 128.0k Stars
LicenseMIT
LanguagePython/TypeScript
Version0.6.0
Updated2026-06-02

📖 Overview

The office mode of Open WebUI enables full utilization of self-hosted AI capabilities in everyday work scenarios. It allows documents such as PDF, Word, and TXT to be imported into a local RAG knowledge base, enabling the AI to answer questions, generate summaries, and write reports based on your actual documents. Multi-user collaboration management is supported, with administrators able to exert fine-grained control over model access permissions. All data is processed locally without being uploaded to any third-party servers, making it particularly suitable for enterprise users with data compliance requirements as well as individual users who need to handle sensitive documents.

✨ Features

  • One-click RAG document Q&A (ready to use with drag-and-drop).
  • Online search using 15+ engines
  • Speech recognition / ASR combined with image recognition.
  • Orchestration of AI agent workflows
  • Multi-user support and secure self-hosted deployment in a private network.

Advertisement

🚀 Quick Start

pip install open-webui
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

🔗 Related Tools