How to Replace macOS Spotlight with Dhito for Private, Semantic Search
TL;DR
macOS Spotlight is a fast application launcher but fails when searching files by memory or scanning complex documents. You can replace or supplement Spotlight with Dhito to gain 100% offline, privacy-first semantic search, automatic meeting transcriptions, image contents searching, and local document QA.
For launching applications or doing quick math, macOS Spotlight (`Cmd + Space`) is deeply ingrained in our muscle memory. It is fast, lightweight, and built directly into the operating system.
However, when it comes to *finding files by memory*, Spotlight shows its age. Because Spotlight relies on keyword-based indexing (lexical matching), it requires you to remember the exact name of a file or the exact phrasing within it. If you search for "hiring guide," but your PDF is named "Recruitment Outline," Spotlight will pass right over it.
Moreover, Spotlight is blind to visual contents inside screenshots, spoken conversation inside meeting recordings, and lacks any ability to let you interact with or summarize your documents.
By setting up Dhito as your primary search brain, you can replace or supplement Spotlight with a local, private AI search engine. Here is a step-by-step guide on why and how to make the switch.
Why Replace or Supplement Spotlight?
Traditional desktop search engines were designed in a pre-AI era. They work by looking up exact letters. If you do not have perfect recall of your filenames, you waste minutes scrolling through files.
Dhito replaces this outdated system with on-device machine learning: 1. Semantic Search: Uses a local vector embedding model (MiniLM) to map the meaning of your files. You can search "financial plans" and find a spreadsheet titled "Q3_Budget.xlsx." 2. Audio & Video Transcription: Automatically runs OpenAI Whisper locally in the background to transcribe meeting files (Zoom, Teams, etc.), making spoken words searchable. 3. Local AI Vision: Uses Microsoft Florence-2 to analyze screenshots, design files, and photos. You can search "invoice whiteboard photo" to find unstructured image documents. 4. Offline Document QA: Runs a quantized local LLM (Qwen) on your Neural Engine. You can chat with your PDFs and get instant summaries and citations without uploading sensitive data to the cloud.
---
Step-by-Step: Setting Up Dhito as Your Search Companion
Because macOS Spotlight is hardcoded into the system, you cannot completely delete it. However, you can configure your workflow so that Dhito handles all deep file search, while Spotlight or a launcher (like Raycast or Alfred) handles application launching.
Step 1: Assign a Primary Hotkey to Dhito
To build new muscle memory, you need a quick key combination to trigger Dhito. We recommend setting a hotkey that is just as easy as `Cmd + Space`.
- Option + Space: The most popular choice. It leaves `Cmd + Space` free for Spotlight (to launch apps) while giving you a matching shortcut for Dhito's deep search.
- Cmd + Option + Space: Another great choice that is easy to trigger with your thumb.
Once configured in Dhito's settings, pressing your chosen hotkey will slide open the Dhito search bar instantly.
Step 2: Configure Spotlight to Focus Only on System Search
To prevent Spotlight from cluttering your screen with irrelevant local search results, you can narrow its scope: 1. Go to System Settings > Siri & Spotlight. 2. Under the "Search Results" list, uncheck Documents, Images, and Folders. 3. Keep Applications, Calculations, and System Settings checked.
Now, Spotlight will remain blazing fast for launching apps and system settings, while Dhito takes over the task of indexing and finding your actual knowledge assets.
Step 3: Index Your Knowledge Folders
Unlike Spotlight, which indexes your entire hard drive by default (including thousands of hidden cache and system files), Dhito allows you to choose exactly which folders contain your work. * Open Dhito and add folders like `Documents`, `Downloads`, `Projects`, or `Screenshots`. * Dhito will build a local vector space of these folders. All files stay 100% on your machine, and nothing is ever sent to the cloud.
---
Comparison: A Day in the Life of a Search Upgrade
| Use Case | macOS Spotlight | Dhito AI Search | | :--- | :--- | :--- | | Search by synonym | Returns 0 results if keywords do not match exactly. | Understands meaning (e.g., "spending" matches "Q2_Expenses.csv"). | | Searching screenshots | Blind to visual elements unless manually tagged. | Scans content (e.g., "analytics dashboard screenshot" finds the image). | | Zoom/Meeting videos | Can only search the video filename. | Transcribes audio locally and matches search terms to spoken timestamps. | | Extracting info | Opens the file, forcing you to read or press `Cmd + F`. | Safe local chat with the PDF using an on-device LLM to summarize. | | Data Privacy | Sends search queries to Apple servers by default. | 100% local, offline, and secure. Zero tracking telemetry. |
The Verdict: Transitioning to the Future of Mac Search
Spotlight is still a useful tool for opening Safari, launching Slack, or doing quick calculations. But for finding and interacting with your files, it is obsolete.
By supplementing your workspace launcher with Dhito, you get the best of both worlds: a fast application shortcut and a private, highly intelligent local search brain that actually understands what your files mean.
Related Articles
Dhito vs Raycast: Which Is Right for You?
Wondering whether to use Raycast or Dhito? We compare Raycast's app launching and automation with Dhito's local semantic search and private document QA.
7 Things Spotlight Can't Find (That Dhito Can)
Spotlight is fast, but it is blind to video speech, image content, and concept meanings. Discover 7 critical things Spotlight fails to find, and how local AI search solves it.
How to Search Inside PDFs on Mac Without Adobe
Looking for an alternative to Adobe Acrobat for searching your PDF collection? Learn the best ways to search inside PDFs on Mac using native tools and local AI.
Want to try Dhito?
Download Dhito and experience the power of local semantic search today.