The voice-to-text shortcut on a Mac: which key, and why it matters
The shortcut that turns your voice into typed text — not the one that reads text back to you. Which key you use, and whether you hold it or tap it, shapes dictation more than the engine behind it does.
Looking for the opposite direction?
This guide is about speech to text — you talk, your Mac types. If what you want is text to speech, your Mac reading a document out loud to you, that is a separate feature called Spoken Content, in System Settings under Accessibility, where you can also set a shortcut that speaks whatever is selected.
Almost everything written about dictation shortcuts stops at where the setting is. That is the easy half, and it takes one sentence: the shortcut lives in System Settings under Keyboard, in the Dictation section, in a menu that lets you pick from a short list of key combinations or set your own.
The half nobody covers is which key to put there and what kind of shortcut it should be — because those two decisions determine whether you dictate every day or give up in a fortnight. A key that another app has already claimed makes dictation feel broken. A shortcut you tap to start and tap again to stop makes dictation feel like something you have to manage. Neither problem is about recognition quality, and no amount of accuracy fixes either one.
Four kinds of dictation shortcut, and what each one feels like
These are not four settings in the same menu. They are four different models of what a shortcut is for, and the difference between them is the whole subject of this page.
| Option | Best for | Where it lives | Where it stops |
|---|---|---|---|
| Press-twice toggle | The built-in Mac dictation, out of the box, with nothing installed | System Settings → Keyboard → Dictation, in the Shortcut menu | Dictation becomes a state you can lose track of: you start talking before it is ready, or leave it running after you have finished. |
| Hold to talk | Short, frequent bursts — a reply, a message, a note, dozens of times a day | Not a built-in option; it comes from third-party dictation apps | Holding a key for twenty minutes is genuinely uncomfortable, so it is the wrong model for dictating a chapter. |
| A spoken command | Hands-free use, where reaching for a key is the thing you cannot do | Voice Control, in System Settings → Accessibility | Something has to be listening the whole time for the command to be heard, which is a bigger commitment than a keypress. |
| An in-app button | One app you write in for hours, where reaching for the mouse once is fine | Inside the app — Word's Dictate button, Google Docs' Voice typing | There is no shortcut to learn, because there is nothing to carry: it does not exist in the next window. |
Table scrolls sideways on a narrow screen.
What your Mac has bound right now
Open System Settings, go to Keyboard, and look at the Dictation section. There is a Shortcut menu there, and whatever it currently says is your answer — that is the key or key combination that starts dictation on this particular Mac.
It is worth checking rather than assuming, because the default has not been the same thing across releases and hardware. Macs have shipped with the function key as the trigger, with a modifier key pressed twice, and — on keyboards that have one — with a dedicated microphone key in the function row wired directly to the same feature. Any article that tells you flatly that “the Mac dictation shortcut is X” is describing one particular year's default.
The same menu lets you choose something else, including a custom combination of your own. Setting it takes about five seconds, so the rest of this page is about what to set it to.
Press-twice to toggle, or hold to talk
This is the decision that matters, and it is not a preference — the two models produce different writing.
A toggle turns dictation into a state. You tap the key, then start speaking, and the first few words are frequently lost because the microphone opened a beat after you did. When you finish you have to remember to stop it, and if you do not, it either keeps listening at an empty room or types a fragment of whatever you said next to a colleague. Both of those failures are small; they happen five times a day, and five small failures a day is how a habit dies.
Hold to talk borrows the model from a walkie-talkie, and it removes the state entirely. The key is the microphone: it is open while you are holding it and closed the moment you let go. You cannot start speaking too early, because you are already holding the key when you begin. You cannot forget to stop, because stopping is letting go of something. There is nothing running in the background between utterances, so there is nothing to check.
It also gives the transcriber something the toggle does not: a clean start and a clean end. When the system knows exactly which stretch of audio is one thought, it can punctuate it as one thought — put a full stop at the end, break a paragraph where you paused, and drop the throat-clearing at the front. A toggle that has been running for four minutes has to guess where your sentences were.
The honest trade-off runs the other way for long-form work. If you dictate in twenty-minute stretches — a chapter, a report, a transcript of your own thinking — holding a key down for the duration is genuinely unpleasant, and a toggle is the better shape for that job. Most people are not doing that. Most people are writing a two-line reply, forty times a day.
Choosing a key nothing else wants
The best dictation key has three properties: your hand can reach it without leaving the keyboard, it does not type a character, and no app has already claimed it.
That last one rules out more than people expect. Letter and number combinations are where every app puts its own shortcuts, so anything built on those will eventually collide with something you care about. The function-key row is contested twice over — once by the media controls printed on the keys, and again by apps that bind F-keys for their own use.
- Modifier keys are the strongest candidates precisely because they carry no character of their own. Right Command and Right Option are, on most people's machines, the two quietest keys on the keyboard.
- The globe key is more contested than it looks. macOS lets you assign it to switch input sources or open the emoji picker, so on a bilingual machine it is often already busy.
- Caps Lock is tempting and frequently already remapped — to an input-source switch, or to Control by anyone who has spent time in a terminal. Check before you take it.
- If you use a combination rather than a single key, keep it to modifiers you can press with one hand. A shortcut that needs both hands defeats the point of not having to type.
- Whatever you pick, use it for a week before changing it. Muscle memory is most of the value, and reassigning the key resets it to zero.
When the shortcut does nothing
A dictation shortcut that fires everywhere except in one app is almost never a broken microphone. The usual causes are boring and quick to check:
- Another app registered the same shortcut. Two apps asking for one key combination is a race, and the winner is not always the same one after a restart.
- The app you are in swallows the key before the system sees it. Editors, remote-desktop windows and full-screen games commonly capture the entire keyboard on purpose.
- A third-party dictation app is missing a permission. Anything that watches for a global shortcut and types text into another app needs accessibility and input-monitoring permission in System Settings under Privacy & Security, and macOS revokes these more readily than you would expect after an update.
- Microphone access was never granted, or was granted to a different app than the one you are now using. Also under Privacy & Security.
- The key is bound twice in your own settings — once for dictation and once for something else you set up months ago and forgot.
If dictation fails in every app rather than one, the problem is upstream of the shortcut, and the troubleshooting post is the faster route.
The other speech shortcuts on a Mac
Three different features here answer to the word “shortcut”, and confusing them wastes an afternoon.
- Dictation's own shortcut, in the Keyboard section of System Settings, is the one this page is about: it starts typing what you say.
- Voice Control, under Accessibility, has its own way of starting and stopping listening and a large vocabulary of spoken commands. It is a different feature aimed at operating the machine, not at typing a sentence.
- Spoken Content, also under Accessibility, is the opposite direction: it reads what is on screen out loud, and it has a speak-selection shortcut of its own. If that is what you were looking for, that is where it is.
- In-app dictate buttons — Word, Google Docs and others — usually have no system shortcut at all, and the ones that do only work inside that app's own window.
What a dedicated shortcut is actually worth
When one key does the same thing in every window, dictation stops being a feature you decide to use and becomes a way of typing. That sounds like a small distinction and it is the entire difference between the people who dictate daily and the people who tried it once.
The threshold is roughly this: the moment you no longer have to think about whether the current app supports it, whether it is already listening, or whether you remembered to stop it, the cost of speaking a sentence drops below the cost of typing it — and at that point you start speaking the sentences you would previously have shortened.
How VoiceSnap Pro handles the shortcut
VoiceSnap Pro is a hold-to-talk app by design. You pick the key, you hold it while you speak, and the text lands at the cursor in whatever app you were already in — the same key, the same behaviour, in every window on the machine. There is no listening state to forget about, because the key is the state.
That is a real trade-off and it is worth naming: if you dictate in long unbroken stretches, holding a key down for twenty minutes is not what you want, and a toggle-based tool is a better fit for that job. It has also not shipped yet — the download page is a waitlist, not a binary — and when it does it is a one-time purchase rather than a subscription.
What it does differently
- One held shortcut that behaves the same way in every app on the machine — no per-app engine, no browser extension, no window to copy out of.
- Punctuation, capitals and paragraph breaks inferred from how you speak, so you never say the word “comma” out loud.
- Filler words and false starts removed on the way in, rather than cleaned up afterwards.
- A custom vocabulary for the names, acronyms and jargon a general model keeps guessing at.
- Every dictation saved to a searchable notes library instead of vanishing into the field it landed in.
- A one-time purchase. No subscription, and nothing you dictate is used to train AI models.
None of this has shipped. There is no download, no trial, no customers and no reviews — the download page collects an email address and nothing else.
Frequently asked questions
Whatever the Shortcut menu in the Dictation section of Keyboard settings currently says, which is the only reliable answer — the default has changed across macOS versions and differs on keyboards that have a dedicated microphone key. Check it there, and change it from the same menu if you want something else.
Yes. The same Shortcut menu offers a short list of key combinations and an option to set a custom one of your own. It takes a few seconds, and it is worth doing if the default collides with something you already use.
No. Apple's shortcut starts a dictation session rather than opening the microphone for as long as a key is held, so you tap to begin and it ends when you stop it or when it decides you have finished. Hold-to-talk behaviour comes from third-party dictation apps.
In order of likelihood: another app has claimed the same key combination, the app you are typing in captures the keyboard before the system sees it, a permission was never granted or was revoked by an update, or the shortcut is bound twice in your own settings. If it fails in every app rather than one, the cause is upstream of the shortcut.
The built-in one works in any standard text field, because it types at the cursor rather than integrating with the app. Apps that draw their own text engine — some code editors, some games, remote-desktop sessions — can intercept the key or handle the inserted text oddly. That limitation applies to every dictation tool, not only Apple's.
On Mac keyboards that have one, the microphone key in the function row triggers the same built-in dictation feature as the shortcut. It is a convenience, not a separate system — and if dictation is switched off in settings, the key has nothing to start.
Related reading
The step-by-step articles, the comparisons and the free tools that go with this guide.
Get told once, on release day
VoiceSnap Pro is a one-time purchase with no subscription, and it is not out yet. One email when it is — nothing else.
One launch email. No newsletter, no sharing your address. Privacy policy