monkeSearch review
A macOS desktop search tool for private, metadata-based file discovery.
Reviewed by iTechGuides Editors · Editorial team · Updated Oct 2026
monkeSearch is an open-source desktop search tool for people who want to find local files on macOS using natural-language queries. It interprets file types, temporal expressions, and keywords, then uses macOS Spotlight to identify matching files. The current approach is read-only and operates on local metadata, making it suited to private file discovery without cloud API calls. Local language-model inference supports the query experience while keeping the search process offline.
Its standout feature is the combination of natural-language parsing and Spotlight predicate search. Users can express searches through file types and time-related terms rather than relying only on conventional keyword matching. The product also integrates with llama.cpp for local inference. This keeps the tool focused on macOS’s existing file-search infrastructure while adding a language-model layer for interpreting queries. The published project also documents a vector-database branch intended for cross-platform experimentation across macOS, Linux, and Windows.
The main boundary is the current search scope: the primary implementation is macOS-only and metadata-only. It does not currently provide file-content search, so it is better suited to locating files by attributes and metadata than finding text inside documents. The vector-database branch broadens the intended platform direction, but it is described as experimental work rather than the main implementation. Choose monkeSearch if offline operation, local inference, Spotlight integration, and natural-language file queries matter most. Choose another tool if you need content search or a current implementation for Linux or Windows.
monkeSearch pros and cons
- Where it wins
- Natural-language queries parse file types and temporal expressions.
- Runs locally without cloud API calls or sending files away.
- Uses macOS Spotlight for read-only local file discovery.
- Where it doesn't
- The current main implementation targets macOS only.
- Search is metadata-only; file-content search is planned.
- The cross-platform vector-database work is experimental.
monkeSearch fact sheet, pricing and score →
Advertiser disclosure: iTechGuides is reader-supported. Vendors can pay for top positions in our rankings and for a place on other products' pages, and we may earn a commission when you click some links. Bought places stand at the number bought; this page and our disclosure say so. Paying vendors may get a fuller, more detailed review from our editors, and that review can change their score. How we rank.
Last updated · How we research and update