Voice to text, locally — macOS menu bar app for private, offline voice dictation.

Download Dictto.dmg

• 100% local processing — All transcription happens on your device. No cloud, complete privacy.
• Menu bar app — No dock icon, always accessible from the menu bar
• Global hotkey — Press Shift+Option+D to start/stop recording
• Multi-language support — Korean, English, Japanese, Chinese, and more
• Direct text insertion — Pastes transcription into active text field (with Accessibility permission)
• Clipboard fallback — Always copies to clipboard
• Custom dictionaries — Add specialized vocabulary for better accuracy
• Sound feedback — Audio cues for recording start/stop
• Input device selection — Choose your microphone from the menu bar or Settings

• Talk to Claude Code — Press Shift+Option+S to speak directly to Claude Code
• Multi-turn conversations — Have back-and-forth voice conversations
• Interactive Q&A — Claude can ask clarifying questions, you respond by clicking options

• Import audio/video files — Supports m4a, wav, mp3, mp4, mov, and more
• SRT subtitle export — Export transcriptions as subtitle files
• Share Extension — Share audio from Voice Memos directly to Dictto
• Progress display — Real-time progress for longer files

• Capture from any app — Transcribe audio playing in Chrome, Safari, or any app in real-time
• Live transcription — See text appear as audio plays
• SRT output — View transcription as subtitles with timestamps
• Time offset correction — Align timestamps to the original video position

• MCP integration — Connect Claude Code via dictto-mcp
• Custom dictionaries — Build vocabulary through natural conversation

• Guided setup — First-launch walkthrough helps you get started
• Customizable shortcuts — Set any key combination for recording
• Automatic updates — In-app updates via Sparkle

One-time purchase. Activate on up to 3 devices.

• macOS 14.0 (Sonoma) or later
• Apple Silicon Mac (M1/M2/M3/M4)

brew tap jhlee111/dictto
brew install --cask dictto

1. Download Dictto.dmg from Releases
2. Open the DMG and drag Dictto to Applications
3. Launch Dictto from Applications
4. Grant Microphone permission when prompted
5. (Optional) Grant Accessibility permission for direct text insertion

1. Launch the app (appears in menu bar)
2. Wait for model to load (~30-40 sec for Large v3 Turbo)
3. Press Shift+Option+D to start recording
4. Speak
5. Press Shift+Option+D again to stop and transcribe
6. Text is automatically pasted (or copied to clipboard)

1. Press Shift+Option+S to open the Agent overlay
2. Speak your request
3. Press Shift+Option+S again to send to Claude Code
4. Wait for Claude's response (streams in real-time)
5. Press Escape or click Close to end the session

Note: Requires Claude Code with dictto-mcp configured. Personal tier required.

1. Click menu bar icon → "Import Audio File..."
2. Select an audio or video file
3. Choose transcription language (optional)
4. Click "Transcribe"
5. Copy the result or export as SRT

1. Click menu bar icon → "Capture audio"
2. Select the app to capture (e.g., Chrome, Safari)
3. Grant Screen Recording permission if prompted
4. Audio is transcribed in real-time as it plays
5. Click Stop to finalize — switch between Text and SRT views
6. Optionally enter a time offset to align SRT timestamps to the video

1. Open Voice Memos app
2. Right-click a recording → Share → "Transcribe with Dictto"
3. Dictto opens and starts transcription automatically

Models are downloaded automatically on first use.

v1.3.0