minirag-mcp is a Python-based local RAG server and CLI that indexes and searches user documents using semantic retrieval combined with BM25 keyword boosting. It supports MCP tools, multilingual embeddings, LanceDB storage, and ingestion of multiple document formats while keeping data on the user's machine.
In the RAG, search & retrieval space, minirag-mcp takes a focused approach. It focuses on searching and retrieving relevant information from personal documents without sending them to a cloud service. It is built as an open-source project for developers and privacy-conscious users building local document search workflows. The project is open source (Open Source). It runs on the web, the command line, and API, and it can be self-hosted.
Behind minirag-mcp is sfrangulov. Among its 9 catalogued features are Hybrid Search, Semantic Search, and BM25 Boosting. It exposes integrations via an MCP server.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Closest matches by what these projects do