quivr

quivr

by Stan Girard

Open-source AI personal assistant for chatting with your files and documents locally

Open Source Edge Computing API Web Python Docker Self-hosted
Visit Product
477 upvotes 596 views

About

Quivr is an open-source project that creates a personal AI second brain — allowing users to upload any type of file (PDFs, documents, spreadsheets, code files, audio, video) and then chat with the AI about their content. The AI can answer questions, summarize documents, extract information, and synthesize insights across multiple files simultaneously.

The project emphasizes privacy and control: by running locally, all uploaded files and queries stay on the user's infrastructure. The system uses vector embeddings to index file content semantically, enabling the AI to retrieve the most relevant information from potentially hundreds of documents in response to any question.

Quivr went viral on GitHub shortly after its release in 2023, attracting tens of thousands of stars as users recognized the practical value of a personal, private document intelligence system. The open-source community has since added features including web scraping, YouTube video ingestion, Notion sync, and integrations with various LLM providers.

Product Features

- Upload any file type: PDF, doc, video, audio, code
- Chat with multiple documents simultaneously
- Semantic search across all uploaded content
- Completely self-hosted for privacy
- Support for OpenAI, Anthropic, and local models
- Web UI for easy interaction
- REST API for integration
- Public and private brain sharing
- Web page and YouTube content ingestion
- Docker deployment for easy setup

About the Publisher

Quivr was created by Stan Girard as an open-source project inspired by the concept of a second brain — an external knowledge base that extends human memory and intelligence. The rapid community adoption validated the demand for private, self-hosted document intelligence. Stan Girard has continued developing Quivr with support from the open-source community and has commercialized some aspects while keeping the core project open-source.