Home/Docs/Search

Search

Keep combines full-text search and AI-powered semantic search so you can find saved content by keywords or by meaning. Just start typing in the search bar at the top of your feed.

How it works

Every search runs in hybrid mode by default. Keep searches your titles, URLs, notes, and tags for exact keyword matches, then also finds semantically similar content using AI embeddings. The two result sets are merged and ranked so the most relevant items appear first.

If you search for "machine learning tutorials", you will also find items about "intro to neural networks" or "deep learning guide" even if those exact words are not in your query.

Filter by tag

Type tag: followed by a tag name to narrow results to items with that tag. Use quotes for tags with spaces.

tag:python -- items tagged "python"

tag:"machine learning" -- items tagged "machine learning"

kubernetes tag:devops -- search for "kubernetes" within items tagged "devops"

Tag names autocomplete as you type.

Filter by collection

Type in: followed by a collection name to limit results to a specific collection. Use quotes for names with spaces.

in:reading-list -- items in the "reading-list" collection

in:"side projects" -- items in "side projects"

react hooks in:frontend -- search for "react hooks" within the "frontend" collection

Collection names autocomplete as you type.

Combining filters

You can combine a search query with tag and collection filters in the same search. Filters are extracted from your query automatically.

authentication tag:security in:backend -- search for "authentication" in items tagged "security" within the "backend" collection

Search via API

The search endpoint accepts the same query syntax. Pass your query as the q parameter.

GET /api/items/search?q=your+query

GET /api/items/search?q=tag:python+intro

You can also pass collection, tags, since, and until as separate query parameters for more precise filtering.