Qmd FAQs

Question 1

What is Qmd and what does it do?

Accepted Answer

Qmd is a high-performance, local-first search engine designed specifically for Markdown files. It indexes your local documents and provides advanced search capabilities, including BM25 full-text, vector semantic, and hybrid search, enhanced with LLM query expansion and reranking.

Question 2

What are the typical use cases for Qmd?

Accepted Answer

Qmd is ideal for developers, researchers, and anyone managing extensive local Markdown-based knowledge bases. It can be used for searching personal notes, project documentation, code snippets, or any Markdown content, providing quick and intelligent retrieval enhanced by AI.

Question 3

What kind of search capabilities does Qmd offer?

Accepted Answer

Qmd provides three main search types: BM25 for traditional keyword matching, vector semantic search for conceptual understanding, and a powerful hybrid search that combines both using Reciprocal Rank Fusion (RRF). It further enhances results with LLM query expansion and reranking for deep semantic understanding.

Question 4

How does Qmd integrate with AI agents?

Accepted Answer

Qmd features an MCP (Markdown Content Protocol) Server, which provides a JSON-RPC 2.0 interface. This allows seamless integration with various AI agents and tools (like Claude Desktop or Cursor), enabling them to leverage Qmd's powerful search capabilities for RAG (Retrieval Augmented Generation) workflows.

Question 5

Is Qmd private and does it rely on cloud services?

Accepted Answer

No, Qmd is built with a private and local-first architecture. It operates entirely on your machine, has no cloud dependencies, and keeps all your data and models completely private. It's distributed as a single binary with minimal runtime requirements.

Qmd

Qmd

주요 기능

사용 사례

주요 기능

사용 사례