Voice-to-text for macOS, right from the menu bar.
Hold a key, speak, release — your words appear wherever your cursor is.

Download the latest release

1. Hold your hotkey (default: Right Command)
2. Speak — the app records from your mic
3. Release — transcription is injected at the cursor

Works in any app: browsers, editors, terminals, Slack, email — everywhere you type.

1. Download Whisper-Smart-mac.dmg from Releases
2. Open the DMG and drag Whisper Smart into Applications
3. Launch — it appears as a mic icon in your menu bar
4. Grant Accessibility and Microphone permissions when prompted

Requires macOS 14 (Sonoma) or later.

Choose the provider that fits your workflow in Settings > Provider:

If the selected provider isn't ready yet, the app falls back to Apple Speech (built-in, zero setup) so you're never stuck.

• Press-and-hold or one-shot — hold the hotkey while speaking, or click "Start Dictation" from the menu for toggle mode
• Left/right key aware — bind specifically to Right Command without Left Command triggering it
• Configurable hotkey — pick a preset or record any modifier combo in Settings
• Terminal-friendly — special paste handling for Terminal.app, iTerm2, Kitty, Warp, Ghostty, and more
• Smart silence detection — if you don't speak, it returns to idle instantly instead of waiting
• Writing styles — neutral, formal, concise, casual, or developer mode with per-app overrides
• Auto-updates — built-in Sparkle updater checks for new versions automatically
• Privacy-first — local providers keep audio on your Mac; cloud is opt-in

Open Settings > Hotkey to:

• Pick a preset: Right Command Hold, Left Control Hold, Option+Space, Control+Space, Fn Hold
• Record a custom combo: Click the shortcut pill and press your key combination
• Left vs right matters: Each physical key is tracked independently

# Build
swift build

# Run
.build/debug/Whisper\ Smart

# Package a DMG
bash scripts/package_dmg.sh

See LICENSE for details.