# qmd

19 Hacker News commenters recommended qmd and 1 criticized it, November 2025 to November 2026. It was recommended in 9 of 13 months. It meets the Hacker Approved badge rule.

**Verdict:** Local Markdown search mixing keyword, vector and LLM reranking, popular as agent memory. Caveat: designed around single-machine setups and downloads several large local models.

## Why people recommend it

- Good, hybrid search results (4 comments): "My agents use both depending on what type of search they are doing, and both provide good results." ([source](https://news.ycombinator.com/item?id=48463395)) "The hybrid BM25 + vector + LLM reranking approach is exactly what I've been wanting for my notes." ([source](https://news.ycombinator.com/item?id=46769465)) "you still get semantic search to help you find the right documents quickly" ([source](https://news.ycombinator.com/item?id=50265332))
- Runs locally with local models (2 comments): "tools like QMD that beautifully download and use local models under the hood" ([source](https://news.ycombinator.com/item?id=48096132)) "I tried to make it not overkill and keep things local." ([source](https://news.ycombinator.com/item?id=46767190))
- Default RAG engine for OpenClaw (2 comments): "now it's used as the defacto RAG engine for OpenClaw" ([source](https://news.ycombinator.com/item?id=47370930)) "it's a great option for OpenClaw!" ([source](https://news.ycombinator.com/item?id=47371356))
- Clever chunking and indexing approach (2 comments): "It uses some cool techniques like resilient chunking and embedding" ([source](https://news.ycombinator.com/item?id=47054673)) "The QMD indexing approach is clever" ([source](https://news.ycombinator.com/item?id=47011347))

## What people warn about

- Built around single-machine setups (1 comments): "qmd is focussed on local to the point of designing around single machine setups" ([source](https://news.ycombinator.com/item?id=48926913))
- Requires several large embedding models (1 comments): "though it requires installing 3 local embedding models (300MB, 640MB, 1.1GB)" ([source](https://news.ycombinator.com/item?id=47539518))

## Cite as

Hacker Approved, "qmd", https://hackerapproved.com/p/qmd/ (counts from 1,795,640 Hacker News comments, November 2025 to November 2026).