home · about · blog · drawings · previous work · lotusvoice · contact · site check · rebuild
LotusVoice
Push-to-talk dictation for macOS. Hold a key, speak, release — your words
appear wherever your cursor is.
Download LotusVoice.dmg
v1.0.0 · macOS only · ~16 MB
How it works
- Hold Right Option — recording starts, menu bar turns red
- Speak — audio is captured via your microphone
- Release — audio is transcribed locally with Whisper AI
- Text is pasted — automatically inserted wherever your cursor is
Everything runs locally on your machine. No audio is sent to the cloud.
Install
- Open
LotusVoice.dmg
- Double-click Install Dependencies — this installs sox (audio recording) and whisper-cpp (transcription) via Homebrew
- Drag LotusVoice.app to your Applications folder
- Launch LotusVoice — grant Accessibility and Microphone permissions when prompted
macOS Gatekeeper notice
LotusVoice is unsigned (no $99/yr Apple Developer account yet). macOS will block it by
default. To run it:
Option 1: Right-click the app → Open → Open
Option 2: Run in Terminal:
xattr -d com.apple.quarantine /Applications/LotusVoice.app
Requirements
- macOS 12+
- Homebrew (the install script handles the rest)
- Accessibility permission (for global hotkey)
- Microphone permission
Features
- Lives in your menu bar — no window clutter
- Transcription history saved as daily markdown files in
~/.voice-history/
- Filters out Whisper hallucinations on silence
- Ignores accidental key taps (under 0.5s)