Omni Search

Search every file on your Mac by meaning. Text, code, PDFs, images, audio, and video in one vector space - searched in any language, entirely on-device.

Features:

• One space for every file: Text, code, PDFs, images, audio, and video are embedded into a single vector space. A text query matches them all - including scanned pages.
• Any language: Languages share that same space, so a query in one finds files in another. Search in English, match notes in German, Chinese, or Japanese.
• Local and private: Indexing and search run on your Mac. Files never leave the device. No accounts, no telemetry, no network at query time.
• Native to macOS: List and Gallery views, real QuickLook thumbnails, drag and drop, and filters by kind, folder, and score.

Private by design. Everything stays on your Mac.

Omni runs a native MLX-Swift port of jina-embeddings-v5-omni in-process, on the Apple silicon GPU. The model downloads once; after that, Omni works offline.

Give your local agents search.

Omni exposes its search over your indexed files as a local endpoint - loopback-only and token-guarded. Agents like Hermes and OpenClaw query your files by meaning, on your machine, with no cloud round-trip.

• Search endpoint: The same semantic search the app uses, over the index you already built - how agents actually reach your files.
• Embeddings · add-on: Raw vectors too, via OpenAI-, Jina-, Cohere-, and Gemini-compatible APIs.

Search with any file, find any file.

Every file - image, audio, video, text - lands in one shared vector space, so a query is just another file. Drop in a photo to find related screenshots, a voice memo to surface where it came from, a PDF to pull up everything near it. Right-click any result and choose Find Similar, or drop a file onto the search. One model, one space, any direction.

How fast is Omni on your Mac?

Real, community-submitted indexing benchmarks over a fixed 1,000-file dataset. Anyone can contribute by running File ? Run Profiling in the app.