Dictation that speaks your language.
TongueType transcribes English, Spanish, French, German, Portuguese, Italian, Japanese, Chinese, Korean, Dutch, and Russian, entirely on your Mac. Pick a language from the menu bar or let it auto-detect. Same hotkey, same price, nothing uploaded.
macOS 14+ · Apple Silicon · Free forever, Pro if you want
The languages
Every one of these is included in the free version and in Pro. There’s no language pack to buy and nothing to download beyond the one Whisper model you already have.
English
The default, and the language Whisper has heard the most of. Accents from everywhere, mumbling included.
Spanish
Español. In our experience both Castilian and Latin American speech transcribe cleanly, accents and all.
French
Français. Liaisons, silent endings, and the accents on the right letters when it writes them down.
German
Deutsch. Compound nouns come out as one word, the way you’d type them.
Portuguese
Português. European and Brazilian speakers both pick the same menu entry.
Italian
Italiano. Fast talkers welcome. Whisper keeps up.
Japanese
日本語. Transcribed into written Japanese, kanji and kana mixed the way you’d type it.
Chinese
中文. Mandarin speech, written out in Chinese characters.
Korean
한국어. Hangul output, spacing included.
Dutch
Nederlands. Yes, the g sounds are fine.
Russian
Русский. Cyrillic output, straight into whatever app you’re in.
Auto-detect
Not sure what’s in a file, or you switch languages a lot? Whisper picks the language from the first few words and runs with it.
Switching is one click
Pick it from the menu bar
Click the TongueType icon and choose under Language. That’s it. The choice applies to live dictation and to file transcription alike, it’s remembered between launches, and if you’ve turned on iCloud sync it follows you to your other Macs. Bilingual at work and at home? Switching back and forth is two clicks a day.
Or let Whisper figure it out
Choose Auto-detect and Whisper identifies the language from the first few words of each recording. It’s handy when you genuinely don’t know what’s in a file, or when you jump between languages so often that picking one is a chore. Two honest caveats: it’s a touch slower than naming the language up front, and on very short phrases it occasionally guesses wrong. If you know what you’re about to say, pick the language explicitly.
Auto-detect also settles on one language per recording. If you switch languages mid-sentence, release the key, switch, and start a new take.
Spoken punctuation in your language
TongueType’s replacement rules turn phrases like “new line” into the symbol you meant. The defaults are English, but the rules are plain text, so adding your own takes a few seconds: “nueva línea” → a line break, “point d’interrogation” → ?, “neue Zeile” → a line break. Matching is case-insensitive and whole-word, and you can keep rules for several languages in the list at once.
The same goes for cancel phrases. Add “olvídalo” or “annule ça” next to “scratch that” and TongueType discards the take when you say it.
Names, places, and the words Whisper gets wrong
Every speech model stumbles on surnames, towns, and jargon, and that’s true in every language. Replacement rules fix the ones you hit most: map the spelling Whisper produces to the one you actually use, once, and it’s handled from then on.
A note on Smart insertion
Smart insertion adds a missing space when your cursor sits right against a word, which is what you want in most languages. If you dictate mainly in Japanese or Chinese, where words aren’t separated by spaces, you may prefer to turn Smart insertion off under Settings → General so the text is inserted exactly as transcribed.
Same app, same privacy, every language
No per-language pricing
Some dictation products sold each language as a separate purchase. Here every language is included, in the free tier and in Pro.
Nothing leaves your Mac
Whisper runs on the Apple Neural Engine. Your voice is transcribed on-device in whatever language you speak, and never sent to a server to be processed or improved upon.
Works abroad, offline
No roaming data, no hotel Wi-Fi, no account that cares which country you’re in. Once the model is downloaded, dictation works anywhere.
Files in any of them
Drop an interview, a voice memo, or a lecture into Transcribe audio/video file… and it’s transcribed in the language you pick, or the one auto-detect finds.
Same hotkey everywhere
Hold, speak, release. The flow doesn’t change with the language, and neither does the overlay, the latch, or the settings.
Nothing to install per language
One Whisper model covers all eleven. Switch languages and you’re dictating immediately, with no extra download.
What it doesn’t do
So you’re not surprised later:
- It doesn’t translate. Speak Spanish, get Spanish. TongueType writes down what you said in the language you said it.
- Auto-detect picks one language per take. For a sentence that starts in English and ends in French, split it into two recordings.
- Eleven languages, for now. Whisper understands more, and we’ve kept the menu to the ones we’ve tested. If yours is missing, tell us. Demand decides what we add next.
Say it in your language.
Download TongueType for free, pick a language from the menu bar, and start talking. Nothing to configure, nothing uploaded.
macOS 14+ · Apple Silicon · Direct download, no App Store