Install the desktop app

This page walks you through installing the IntelliPrompter desktop app on macOS. You should have a beta invite email in hand before you start. Plan on about 3 minutes, plus a one-time model download on first launch.

• macOS 14.2 (Sonoma) or newer.
• Apple Silicon (M1 or later) or Intel.
• 8 GB RAM minimum. 16 GB recommended if you plan to run Ollama locally.
• About 500 MB of disk for the app itself.
• Roughly 145 MB extra per Whisper transcription model you keep installed.
• A beta invite email with your DMG download link.

1. Open your beta invite email.
2. Click the Download for macOS link.
3. Your browser downloads a file named something like IntelliPrompter-0.X.Y-arm64.dmg (or -x64.dmg on Intel). Save it to your Downloads folder.

1. Double-click the DMG to mount it. A Finder window opens.
2. Drag the IntelliPrompter app icon onto the /Applications shortcut in the same window.
3. Wait for the copy to finish. Eject the DMG from Finder.
4. Open the app from /Applications, not from the mounted DMG. Running it from the DMG causes auto-update to fail and permission prompts to re-ask every launch.

The first time you open the app, macOS shows a series of permission prompts. Here is what each one is for and whether you need to say yes.

• Microphone. Required. The app cannot transcribe your voice without it.
• Screen Recording. Required for capturing system audio (the other side of the call) in Full privacy mode. Without it you only get your own voice.
• Accessibility. Optional. Grant it if you want global hotkeys to work while another app has focus (for example, snapshot-bookmark during a Zoom call).

During the beta the installer is signed and notarized, but macOS occasionally still shows a “IntelliPrompter” is damaged or cannot be opened dialog on first launch. If that happens:

1. Click Cancel on the Gatekeeper dialog.
2. Open System Settings.
3. Go to Privacy & Security.
4. Scroll to the bottom. You should see a row that reads “IntelliPrompter” was blocked from use because it is not from an identified developer.
5. Click Open Anyway next to that row.
6. A second dialog appears asking you to confirm. Click Open. You will not see this prompt again.

The first time you sign in, the app downloads a Whisper transcription model so that speech-to-text runs locally on your machine. The default base model is about 145 MB. Expect roughly 30 to 90 seconds on a reasonable wifi connection.

• Be on wifi. The download retries on cellular tethering, but it is slow.
• You see a progress bar in the Setup screen. Do not quit the app during the download.
• Once complete, the model is cached. Future launches are instant and work fully offline for transcription.

• Getting Started. Sign in, add an AI provider, and run your first meeting.
• Bring your own key. How to plug in your own Anthropic, OpenAI, Gemini, or Ollama access.
• Privacy modes. What each of the three modes captures, and which data leaves your machine.