TongueType
Local, privacy-focused Whisper dictation for macOS
Last verified:
What is TongueType?
TongueType is a macOS voice dictation app that lets you speak to your Mac and have your words turn into typed text instantly, using OpenAI’s Whisper model running entirely on your device. It works as an alternate input method: you hold a hotkey, speak, then release, and the transcription appears wherever your cursor is, whether you’re writing emails, Slack messages, code comments, or LLM prompts. The app is designed to be lightweight and disappear from your workflow while still being powerful enough for long-form writing and communication.
A key feature is local, offline transcription: all audio is processed on your Mac using Apple Silicon’s Neural Engine, so your voice never leaves your device and there is no cloud, no account, and no subscription. TongueType also supports dragging and transcribing audio and video files (WAV, MP3, MP4, MOV) so you can quickly turn meeting recordings, interviews, or voice memos into text without uploading anything. The app is tuned for privacy‑minded users who want fast, accurate dictation without sending data to remote servers.
TongueType is aimed at macOS users on Apple Silicon Macs who want a fast, natural way to type, especially those who find keyboards uncomfortable or slow, such as people with RSI, arthritis, carpal tunnel, or other hand‑related conditions. It also appeals to developers, writers, and knowledge workers who want to dictate code comments, commit messages, documentation, or long messages without relying on Apple’s built‑in dictation or web‑based transcription services. The configurability of hotkeys, overlay position, accent colors, and Rainbow Mode lets users tailor the experience to their own style and workflow.
TongueType pricing
Pricing model: Freemium
TongueType is free with all features included, capped at 30 minutes of live dictation per calendar month and 10 seconds of each file for transcription. TongueType Pro costs 19.99 USD as a one‑time purchase, unlocks unlimited live dictation and full‑length file transcription, and is licensed for use on up to 5 Macs. There are no subscriptions or recurring charges.
TongueType pros
- Runs 100% locally on your Mac with no cloud processing
- Uses Whisper AI for high‑quality speech recognition
- No accounts, no email, no personal data collected
- No subscriptions; free tier includes all features
- One‑time Pro payment for unlimited dictation and full‑file transcription
- Hotkey‑based dictation feels fast and frictionless
- Supports 12 common languages plus automatic language detection
- Transcribes audio and video files by dragging them in
- Works in any app where you can place a cursor
- Helpful for people with typing difficulties or chronic pain
- Menu bar‑only design with no Dock icon or main window
- Customizable hotkey, overlay position, and accent colors
- Rainbow Mode for a colorful recording indicator
- Privacy‑first design: no servers or analytics infrastructure
- Lightweight and simple to install with no App Store dependency
TongueType cons
- Requires macOS 14+ and Apple Silicon Macs (no Intel support)
- Free tier limits live dictation to 30 minutes per month
- Free tier caps file transcription at 10 seconds per file
- No iOS version available due to platform restrictions
- Only works on macOS, not Windows or Linux
- Relies on menu bar icon, which can be hidden by other apps
- No built‑in collaborative or team‑licensing model
- No offline‑only language packs; still uses a large local model
Frequently asked questions about TongueType
Is TongueType really free?
Yes. TongueType is free with all features included. The free tier gives you up to 30 minutes of live dictation per calendar month and allows file transcription with only the first 10 seconds of each file processed so you can try it. If you want unlimited dictation and full‑length file transcription, TongueType Pro is a one‑time 19.99 USD purchase that works on up to 5 Macs with no recurring fees.
Does my audio get sent anywhere?
No. TongueType runs Whisper AI locally on your Mac’s Apple Silicon, so all audio is processed on your device. There are no servers, no analytics, and no data collection. Your audio never leaves your Mac, and the developers do not have access to your recordings or transcriptions.
Why does it require Apple Silicon?
TongueType runs the Whisper AI model on Apple’s Neural Engine using CoreML, which only works on M1 and later chips. This lets the app run efficiently and offline, but it also means older Intel‑based Macs are not supported and cannot run TongueType.
Where is TongueType after I install it? I don’t see a window.
TongueType is a menu bar app, so it has no Dock icon and no main window. Look for its icon in the menu bar at the top‑right of your screen, near the clock. If it’s not visible, your menu bar may be full or the icon may be hidden behind the notch or behind other app menus; repositioning windows or quitting some menu bar apps should reveal it. It can also be tucked into a menu bar manager’s hidden section, which you can drag into the visible area.
How is this different from macOS built‑in dictation?
Apple’s built‑in dictation can send audio to Apple servers and must be enabled in System Settings, while TongueType is entirely local and does not require an online account. TongueType adds a configurable hold‑to‑record hotkey, support for audio and video file transcription, smart post‑processing with custom rules, and zero data sent to any cloud service, giving you more control and privacy.
How accurate are the transcriptions?
TongueType uses OpenAI’s Whisper model, which is highly accurate even with accents, background noise, and casual speech. It’s not perfect and may still require light editing, but in practice users typically spend less time correcting transcriptions than they would have spent typing the same text manually.
What languages does it support?
TongueType supports English, Spanish, French, German, Portuguese, Italian, Japanese, Chinese, Korean, Dutch, and Russian, plus automatic language detection so you can switch between languages without reconfiguring the app. This makes it useful for multilingual users who often mix languages in their speech.
Can I use it for coding?
Yes. TongueType works anywhere your cursor can type, including code editors, terminals, and IDEs. You can dictate code comments, commit messages, documentation, configuration notes, and even Slack or GitHub messages about your code. It respects the context of the app you’re in, so transcriptions land exactly where you click or type.
Is TongueType useful for people who have difficulty typing?
Yes. TongueType is designed as an alternate input method for users who struggle with keyboards due to conditions like arthritis, carpal tunnel, tendonitis or RSI, tremor, hemiparesis after stroke, or limb difference. By holding a hotkey, speaking naturally, and releasing, users can get text into their apps without relying on physical typing, making it a practical assistive tool for many.
Is TongueType available for iOS?
Not yet. iOS does not currently allow keyboard extensions to directly access the microphone, so any voice‑typing keyboard would have to bounce you out to a separate app to record and then back again, which creates a clunky and frustrating experience. The developers have not shipped an iOS version because they want to maintain a smooth, high‑quality flow and are waiting for Apple to allow proper microphone access for keyboard extensions before releasing an iPhone version.