QuivrHQ/quivr

Health v2 · maintenance only

Adoption not part of the score

39579 stars · 3739 forks observed · 2026-10-05

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

Quivr is an opinionated RAG (Retrieval-Augmented Generation) library, available as the quivr-core Python package, that lets developers ingest files and ask questions over them with a few lines of code. It works with any LLM (OpenAI, Anthropic, Mistral, Groq, Ollama) and any vector store (PGVector, Faiss), and powers the hosted Quivr.com 'second brain' product.

Use cases

• build a chatbot that answers questions over my pdf documents
• add RAG to my existing app without building retrieval from scratch
• chat with my files locally using Ollama
• ingest markdown and txt files and query them with an LLM
• swap between GPT-4, Mistral, and Groq in a retrieval pipeline
• create a personal 'second brain' knowledge assistant
• customize a RAG workflow with internet search and tools

When to choose

• you want a batteries-included, opinionated RAG pipeline instead of assembling LangChain pieces yourself
• you need flexible LLM and vector-store backends (OpenAI, Anthropic, Mistral, Ollama, PGVector, Faiss)
• you want quick document Q&A with minimal setup, including PDF parsing via Megaparse integration

When to avoid

• you need full low-level control over every retrieval and chunking step
• you want a fully managed SaaS without self-hosting or Python integration
• your project is not Python-based and you cannot use the quivr-core package or its API

Facets

library · maturity active

rag llm-inference chatbot agent-framework nlp pdf vector-database large-language-models artificial-intelligence chatbots developer-tools python self-hosted cross-platform second-brain quivr-core opinionated-rag document-qa megaparse ollama pgvector faiss retrieval-augmented-generation natural-language-processing docker

Member repositories

For agents

markdown · JSON · MCP: product_card(name="The-Vibe-Company/quivr")