Semantic search
pendingby Shreyas
Search vault notes with local semantic retrieval powered by QMD.
Semantic search
Fully-local semantic search for Obsidian notes, powered by QMD.
Why this is useful
Keyword search fails when you do not remember exact wording. Semantic search helps you find the right note by meaning, not only literal terms.
This plugin gives you a native Obsidian sidebar that:
- Finds the closest notes for your query
- Shows the most relevant snippet for each note
- Opens the target note on click
Fully local by design
- Search runs on your machine through your local
qmdcommand. - No cloud search API is required.
- No telemetry is sent by this plugin.
Features
- Sidebar workflow (
Open sidebar) - Search selected text from editor (
Search selected text) - One-click local embedding refresh (
Run embeddings now) - Configurable mode, score threshold, collection filter, and timeout
Attribution
This plugin is a UI wrapper around QMD and relies on QMD for indexing and retrieval:
- QMD repository: https://github.com/tobi/qmd
- QMD npm package: https://www.npmjs.com/package/@tobilu/qmd
Requirements
- Obsidian desktop
- QMD installed locally (
qmdin PATH or full executable path in plugin settings) - Your notes indexed in QMD
Quick setup
- Index your vault in QMD:
qmd collection add /path/to/your/vault --name vault
qmd embed
- Build this plugin:
npm install
npm run build
- Copy
main.js,manifest.json, andstyles.cssto:<Vault>/.obsidian/plugins/semantic-search/ - Enable Semantic search in Obsidian Community plugins.
Usage
- Run
Open sidebar. - Enter a query in natural language.
- Click a result to open the note.
- If results are stale, run
Run embeddings now.
For plugin developers
Search results and similarity scores are powered by semantic analysis of your plugin's README. If your plugin isn't appearing for searches you'd expect, try updating your README to clearly describe your plugin's purpose, features, and use cases.