Voice to text guide: how to use speech to text on any device
A plain voice to text guide: what speech to text is, how to turn on voice typing on Mac, iPhone, Android, Windows and Google Docs, and how to get cleaner results.
Speech to text turns what you say into typed words, and you already have it: every Mac, iPhone, Android phone and Windows PC ships with voice typing built in. On a Mac, turn on Dictation in System Settings › Keyboard, click where you want the words, and press the Dictation shortcut. On Windows press WindowsH; on a phone, tap the microphone on the keyboard.
The rest of this guide covers each device step by step, how to speak so the result needs less fixing, and when a dedicated app is worth it. Steps come from Apple, Google and Microsoft's own support pages, checked November 2026.
What speech to text actually does
A speech recognition model listens to short slices of audio and predicts the most likely words. Older systems needed you to train them on your voice. Current ones work out of the box, add punctuation for you in supported languages, and run either on your device or on a company's servers.
That last part matters more than people think. On-device recognition works offline and your audio stays put. Cloud recognition needs a connection and sends your voice somewhere else to be processed. Our local vs cloud dictation guide explains the trade-off in detail.
People use a lot of names for the same thing: voice typing, dictation, talk to text, type with voice. One distinction is useful, though. Dictation is speaking to write new text right now. Transcription is converting a recording you already have. They need different tools, which dictation vs transcription covers.
How to do speech to text on each device
| Device | Where to turn it on | How to start | Needs internet? |
|---|---|---|---|
| Mac | System Settings › Keyboard › Dictation | Dictation shortcut, or Edit › Start Dictation | Not for many languages on Apple silicon |
| iPhone | Settings › General › Keyboard › Enable Dictation | Tap the mic on the keyboard | Not for many languages |
| Android | Install Gboard | Tap the mic at the top of the keyboard | Depends on phone and language |
| Windows 11 | Built in | Windows key + H | Yes |
| Google Docs | Tools › Voice typing | Click the mic box | Runs through the browser |
Mac
- Choose Apple menu › System Settings, click Keyboard, and turn on Dictation. Click Enable if asked.
- Click into any text field: Mail, Notes, a browser, Word.
- Press the Microphone key if your keyboard has one, use the Dictation shortcut (pressing fn twice is common), or choose Edit › Start Dictation.
- Speak. Press Esc or the shortcut again when you're done.
This works the same on macOS 14 Sonoma and macOS 15 Sequoia. Dictation adds commas, periods and question marks automatically in supported languages, and stops by itself after 30 seconds without speech. To change language, click Edit next to Languages in the same section. The full walkthrough is in how to dictate on a Mac.
iPhone
- Open Settings › General › Keyboard and turn on Enable Dictation.
- Tap in a text field, then tap the microphone on the keyboard.
- Speak, then tap the mic again or stay quiet and it stops after 30 seconds.
The keyboard stays open while you dictate, so you can type a correction and keep talking. Apple says dictation is processed on the iPhone in many languages with no internet needed.
Android
On most Android phones the keyboard is Gboard. Open any app you can type in, tap a text field, tap the microphone at the top of the keyboard, and talk when you see "Speak now". The first time, Android asks for microphone access: choose While using app.
Windows
Put the cursor in a text box and press WindowsH. Wait for "Listening..." and speak. Say "stop listening" or click the mic to stop. Microsoft says it uses online speech recognition, so you need to be connected. Windows 11 can insert punctuation automatically; turn it on from the voice typing settings (the gear icon).
Google Docs
In Chrome, Edge or Safari, open a document and choose Tools › Voice typing. Pick your language in the microphone box, click the mic, and speak. Spoken punctuation and editing commands only work in English.
How to speak so you fix less
Speech to text is fast, but the time you save disappears if you spend it correcting. A few habits help on every device:
- Think one sentence ahead. Most mistakes come from restarting a sentence halfway. Pause before you speak, not during.
- Say punctuation when it matters. Auto-punctuation guesses commas well and colons badly. Say "colon", "new line" or "new paragraph" when the structure matters. Our punctuation commands list has the Mac set.
- Use a decent mic at a steady distance. Laptop mics are fine in a quiet room. In a café, a headset helps more than anything else.
- Correct afterwards, not during. Dictate the whole paragraph, then fix names and numbers in one pass.
- Spell rare names once. On iPhone you can say "Let's invite Steven S-T-E-P-H-E-N" to spell a word.
If voice typing won't start at all, go to voice typing not working for fixes on every platform.
When the built-in tool is enough
For short messages, a search, or a quick note, the built-in options are good and free. Use them. On a Mac with Apple silicon, Apple's Dictation runs on the device and needs nothing extra.
They get frustrating when you write a lot by voice. Apple's Dictation keeps every "um", times out when you pause, spells your colleagues' names differently each time, and gives you the same flat transcript in Slack as in a formal email. Windows and Google Docs voice typing need a connection.
On a Mac: a faster way to type with your voice
Sayburst puts speech to text on a single key. Hold fn, talk, let go, and cleaned-up text is inserted where your cursor is, in any app. Recognition runs on your Mac with Parakeet or Whisper, so it works offline once the model is downloaded, and your audio is never uploaded.
What you say into a Slack thread:
"Hey Maya, I looked at the, uh, the pricing deck and I think slide four needs the new numbers, the Q3 ones, not Q2"
What Sayburst types:
Hey Maya, I looked at the pricing deck and I think slide four needs the new numbers, the Q3 ones, not Q2
Fillers and the repeated "the" are gone, and with Slack set to Message mode there's no forced final period. The same key in Mail gives you an email with a greeting and paragraphs. It costs $29 once for one Mac after a free 7-day trial; there's no subscription. Want to compare it with other options first? The best talk to text apps lists what works on every device.
Try Sayburst free for 7 days: hold fn, talk, and clean text lands in any Mac app.
Download SayburstQuick answers to common setup problems
- No mic button on iPhone: Enable Dictation is off.
- Mac shortcut does nothing: check Shortcut and Microphone source under Keyboard › Dictation, and make sure Voice Control is off.
- Wrong language: on a Mac, add it under Dictation › Languages; on a phone, switch keyboards.
- Windows says it can't connect: voice typing needs the internet; check your connection.
Speech to text is one of those tools that feels awkward for a day and then saves you an hour a week. Start with what's built in, learn to speak a sentence at a time, and move to a dedicated app when you notice how often you're cleaning up after it.
Hold fn, talk, and clean text lands where your cursor is. Free for 7 days.
Download SayburstQuestions people ask
What is speech to text?
Speech to text is software that listens to your voice and writes the words out as text. Every modern phone and computer has it built in, usually as a microphone button on the keyboard or a keyboard shortcut.
How do I do speech to text on a Mac?
Turn on Dictation in System Settings › Keyboard, click where you want to type, then press the Dictation shortcut or choose Edit › Start Dictation. Speak, then press Esc or the shortcut again to stop.
How do I use speech to text on Windows?
Put your cursor in a text box and press Windows key + H. Wait for "Listening..." and speak. Windows voice typing needs an internet connection.
Is voice typing the same as speech to text?
Yes, for everyday purposes. Voice typing, dictation, talk to text and speech to text all describe speaking and getting typed words. Transcription usually means converting a recording you already have.