Lispr ← All posts
Guides

How to use voice typing on Windows

July 4, 2026 · 10 min read

Every Windows PC can type from your voice. The feature ships with Windows, costs nothing, and lives one shortcut away: Win+H. Microsoft keeps it quiet, though, so plenty of people never find it.

This guide covers how to use voice typing on Windows from start to finish: the shortcut, the spoken commands, the language rules, and how it differs from voice access, the other speech feature built into Windows 11. It also marks the points where voice typing tends to give up, so you can tell whether the built-in tool is enough or whether you want something faster.

What voice typing on Windows actually is

Voice typing is software that listens and writes down what you say. On Windows you have three routes. The first is voice typing itself, the Win+H feature, built in and free. The second is voice access, a Windows 11 accessibility feature that controls the whole PC by voice and can dictate too. The third is a dedicated app that adds speed, automatic language handling, or a cleaner way to start and stop.

Start with the tool already on your PC before you install anything.

How to start voice typing (Win+H)

Voice typing needs three things: an internet connection, a working microphone, and your cursor in a text box. Microsoft's speech recognition runs in the cloud — powered by Azure Speech services — so there is no offline mode for the recognition itself.

  1. Click into the field where you want the words — a document, a chat box, a browser bar.
  2. Press Win+H. On a touch keyboard, press the microphone button instead.
  3. Wait for the "Listening..." alert.
  4. Speak at a steady, natural pace.
  5. Say "stop dictating" or press the microphone button when you are done.

The text lands in the active text field. That "active text field" part matters: voice typing is not a system-wide transcriber, and if your cursor is not in a box that accepts text, there is nowhere for the words to go.

On Windows 10

Win+H works on Windows 10 too, with a catch: it depends on the Online speech recognition setting (Start > Settings > Privacy > Speech), and it is far more English-centric — basic dictation covers about a dozen locales across seven languages, and the editing commands work in US English only. Note that Windows 10 reached end of support in October 2025, so the built-in tools there are frozen where they stand.

Voice typing commands

Voice typing will not guess every mark for you by default. You say punctuation out loud as part of the sentence:

If you would rather not speak the marks, open the gear icon on the voice typing flyout and switch on Auto punctuation, and Windows will insert them from your phrasing.

Editing commands work too, in the supported languages: say "delete that" to remove the last phrase, "select that" to highlight it, or "press Enter" to send. To stop, "stop dictating", "stop listening" and a handful of variants all work. The same gear menu holds a profanity filter, a microphone picker, and the voice typing launcher, which pops the flyout up automatically whenever your cursor lands in a text box.

For cleaner punctuation habits that carry across every tool, see how to dictate punctuation cleanly.

Which languages voice typing supports

Windows 11 voice typing supports 46 locales covering about 36 languages — the majors are all there: English in six flavors, Spanish, Portuguese, German, French, Italian, Polish, Russian, Chinese, Japanese, Korean, Hindi, Turkish, Vietnamese, Thai, and most of the EU set.

Two rules to know:

The dictation language follows your input language. There is no automatic detection. To dictate in another language, you switch the Windows input language first — the language switcher in the taskbar corner, or Win+Spacebar — and then start voice typing again. If you dictate in two languages within one message, that is a settings round-trip every time you switch.

If your input language is not on the list, voice typing refuses. You get "Voice typing isn't available in the current language" and a pointer to install a supported one. The absences are not exotic: Ukrainian, Indonesian, Arabic, Hebrew, and Greek are all missing from the list.

Voice typing vs voice access

Windows 11 ships a second speech feature, and the two are easy to mix up:

The trade is language coverage: voice access speaks about seven languages (fifteen dialect variants), against voice typing's thirty-six. If you need offline dictation in English, Spanish, German, or French, voice access is the built-in answer. For most other languages, it is the cloud path or nothing.

On newer Copilot+ PCs there is also fluid dictation, an on-device layer that cleans up grammar, punctuation, and filler words as you speak — English only for now, and the speech recognition underneath still runs online.

Privacy: where your voice goes

With online speech recognition, your audio is sent to Microsoft's speech services to be transcribed. Microsoft's wording: voice data is sent "only to provide the service and create text transcriptions", and is not stored, sampled, or listened to without permission. The switch lives at Settings > Privacy & security > Speech, and turning online recognition off also turns voice typing off — the on-device option there covers voice access, not Win+H.

Where voice typing stops

Voice typing is solid and free, and for a short message in one language it is often all you need. Three limits show up once you lean on it for real work.

One language at a time, by hand

No detection, no mid-sentence switching. Every language change is a trip through the input switcher, and if your language is not on the 36, the feature is simply unavailable. For anyone who writes in two languages a day — or whose language did not make the list — this is the wall.

It types what you said, nowhere else

Voice typing has no translation. You dictate in language X and get text in language X; getting it into another language is a separate copy-paste trip through a translator.

It is a toggle, not push-to-talk

You press Win+H, a flyout opens, you wait for "Listening...", and the microphone stays open until you stop it. There is no mode where you hold a key while you speak and release to finish — the control that removes any doubt about when the microphone is on. Push-to-talk vs always-listening dictation covers why that difference matters.

None of this makes voice typing a bad tool. It is free, already installed, and the editing commands are genuinely good. For a one-language, one-message workflow it holds up.

The faster path: Lispr

When the language dance or the missing translation gets in the way, a dedicated tool fixes the exact points above. Lispr for Windows is a small tray app built for this. You hold the right Ctrl key, talk, and let go — the text appears at your cursor in whatever app you are in. No flyout, no waiting on an indicator.

Voice typing (Win+H) Lispr
Activation Toggle, wait for "Listening..." Hold a key, release to stop
Languages 46 locales, switched by hand ~99, detected for you
Translation None Built in — press a second key
Mid-sentence language switch Settings round-trip Automatic
Account None None
Price Free, built in Free, no account
Internet Required Required
Runs on Windows 10/11 Windows 10+, 64-bit, ~8 MB installer

A few things set it apart from the built-in option:

Like voice typing, it runs in the cloud, so it needs a connection — audio is transcribed and then discarded, nothing is stored. If you want the full story, the Windows launch post covers what shipped, and the download page has both platforms.

Closing thoughts

Using voice typing on Windows comes down to one shortcut and a handful of spoken commands: put the cursor where the words should go, press Win+H, and talk. Switch on Auto punctuation, learn "delete that", and you can dictate a surprising share of your day.

Use it for a week. If the manual language switching, the missing translation, or the toggle-style microphone starts to cost you, that is the point to try a dedicated app. There is no reason to install anything before the free tool has shown you its edges.

FAQ

How do I turn on voice typing in Windows 11?

There is nothing to enable in advance. Click into any text field and press Win+H (or the microphone button on the touch keyboard), wait for the "Listening..." alert, and speak. You need an internet connection and a working microphone.

Does Windows voice typing work offline?

No. Voice typing uses online speech recognition powered by Azure Speech services, so it needs an internet connection. The offline alternative built into Windows 11 (22H2 and later) is voice access, which runs on-device but supports far fewer languages.

How do I change the voice typing language?

Voice typing follows your Windows input language. Switch it with the language switcher in the taskbar corner or with Win+Spacebar, then start voice typing again. There is no automatic language detection.

What languages does Windows voice typing support?

Windows 11 voice typing supports 46 locales across roughly 36 languages, including English, Spanish, Portuguese, German, French, Italian, Polish, Russian, Chinese, Japanese, Korean, Hindi, Turkish, Vietnamese, and Thai. Ukrainian, Indonesian, Arabic, Hebrew, and Greek are not on the list.

What is the difference between voice typing and voice access?

Voice typing (Win+H) dictates into text fields and runs in the cloud. Voice access, available on Windows 11 22H2 and later, controls the entire PC by voice and dictates too, and it works offline on-device — but it covers only about seven languages against voice typing's thirty-six.

Why does it say "Voice typing isn't available in the current language"?

Your current input language is not one voice typing supports. Install and switch to a supported input language (Win+Spacebar switches between installed ones), or use a dictation tool with automatic language detection.

Can Windows voice typing translate what I say?

No. Voice typing types exactly what you said, in the language you said it. If you want to speak one language and type another, you need a separate tool — Lispr does this with a second held key, in one gesture.

Try Lispr

Voice to text in any Mac app — hold a key, talk, let go. Free, no account, ~4 MB.

Download for macOS