QAnything

RAG & Memory 💻 Python ⚖️ AGPL-3.0 🟢 Actively maintained
14.1k stars

QAnything is a local knowledge base question-answering system that supports a wide range of file formats and databases, allowing for offline installation and use. It solves the problem of retrieving accurate answers from large-scale local documents by using a two-stage retrieval process. It is designed for developers and businesses that need a secure, on-premises Q&A solution.

✨ Key features

  • Supports multiple file formats including PDF, Word, PPT, Excel, Markdown, Email, TXT, Image, CSV, and HTML.
  • Two-stage retrieval (embedding + reranking) for improved accuracy with large data volumes.
  • Data security: can be installed and used fully offline.
  • Cross-language Q&A: supports Chinese and English regardless of file language.
  • Hardware friendly: runs on CPU-only environments, supports Windows, Mac, Linux.
  • One-click installation and deployment with Docker Compose.

🎯 Use cases

  • Build a local knowledge base for internal documentation search.
  • Provide accurate answers from large collections of PDFs, Word docs, and other files.
  • Enable offline Q&A for sensitive data that cannot leave the premises.
  • Create custom bots for specific domains using retrieval and generation.
  • Use as a backend for RAG applications with support for OpenAI-compatible APIs.

📦 Installation

🧰 Requirements: Requires Docker and Docker Compose; supports Windows, Mac, and Linux. No GPU required for version 2.0.0.

step1: pull qanything repository
git clone https://github.com/netease-youdao/QAnything.git
step2: Enter the project root directory and execute the startup command.
cd QAnything
docker compose up -d

🚀 Usage

# After installation, open the web interface at http://localhost:8777
# Upload your documents and ask questions.

⚠️ Good to know

Version 2.0.0 no longer provides a default local LLM; all LLM access is through the OpenAI interface, and users need to deploy a local LLM via tools like Ollama.

❓ FAQ

What file formats does QAnything support?

It supports PDF, Word (docx), PPT (pptx), Excel (xlsx), Markdown (md), Email (eml), TXT, Image (jpg, jpeg, png), CSV, and Web links (html).

Can QAnything run offline?

Yes, it supports installation and use by unplugging the network cable throughout the process, ensuring data security.

Does QAnything require a GPU?

No, version 2.0.0 defaults to running in a pure CPU environment and supports Windows, Mac, and Linux.

How do I start QAnything?

Clone the repository, run docker compose up -d in the project root, and then access http://localhost:8777.

📊 Repository

Stars★ 14,086
Forks🍴 1,346
Open issues🐛 405
Last commit🕒 Sep 3, 2026
Created📅 Jan 2024
Language💻 Python
License⚖️ AGPL-3.0

🤖 Overview, features, install steps and FAQ were generated from the project's README on Sep 4, 2026. Always check the original source before running commands.