User Guide

System Requirements

• macOS 15 (Sequoia) or later
• Apple Silicon — M1, M2, M3, or M4 series
• Microphone (built-in or external)

Installation

Download Airgap Voice from the Mac App Store, or request an Evaluation Kit to receive a direct download link.

First Launch

Airgap Voice is a menu bar utility. After installation, click the Airgap Voice icon in your Mac's menu bar to open the main menu. There is no Dock icon and no main window — the menu bar is your primary interface.

Why Is This Needed?

Airgap Voice uses macOS Accessibility to type transcribed text directly into the active application via simulated keystrokes. Without this permission, transcription will work but the text cannot be inserted into other apps.

Step-by-Step

On first launch, the app presents the Accessibility Permission Required dialog:

1. Click Open Accessibility Settings.
2. In System Settings → Privacy & Security → Accessibility, find Airgap Voice.
3. Toggle the switch to ON.

Click the Airgap Voice icon in the menu bar to reveal the main menu:

How It Works

1. Click Start Recording in the menu or press ⌥⌘R.
2. Speak naturally — transcribed text appears in real time (< 250 ms latency) in whichever application currently has focus.
3. To stop, click Stop Recording, press ⌥⌘R again, or let Auto-Stop handle it (if enabled in Transcription Settings).

100% Local Processing

All audio processing runs entirely on your Mac's GPU using a neural network. No audio data ever leaves your device — the engine is physically incapable of connecting to the internet. You can verify this with Activity Monitor or a network inspector like Little Snitch.

Streaming Architecture

Unlike batch-based dictation tools that make you wait for a text dump, Airgap Voice uses 80 ms streaming chunks. Words appear on screen as you speak, preserving your flow state and cognitive rhythm.

Hover over Microphone in the main menu to see all detected audio input devices:

• System Default — Uses whatever macOS has set as the default input device.
• Named Devices — Select a specific microphone. The currently active device is indicated with a checkmark (✓).

Open the Settings window from the main menu → Settings:

Shortcuts

Toggle Recording — The global keyboard shortcut to start/stop recording. Default is ⌥⌘R. Click the recorder field to set a custom shortcut.

Behavior

Confirm before quitting — When enabled, the app asks for confirmation before quitting to prevent accidental closure during a recording session.

Navigate to the Transcription tab in the Settings window:

Auto-Stop Recording

When Enable Auto-Stop is toggled on, recording stops automatically after a period of silence. The silence duration is configurable from 7 to 60 seconds.

This is useful for hands-free dictation: simply stop speaking and the recording ends on its own.

Filler Words

When Enable Filler Words Removal is toggled on, common filler words are automatically stripped from the transcription output. The default list includes:

um, uh, hm, hmm

You can customize this list by editing the comma-separated text field.

The Support tab provides quick access to help and service channels:

Contact

Service

Links to the Privacy Policy and Terms of Service are also available at the bottom of this panel.

The About panel displays:

• The Airgap Voice app icon
• Version number and build hash
• Copyright notice
• Links to Privacy Policy and Terms of Service