It depends on your platform. On Windows 11, press Win+H in any text field. On a Mac, enable Dictation in System Settings → Keyboard, then press the shortcut key (F5 or the microphone key by default). In Google Docs, open Tools → Voice typing or press Ctrl+Shift+S (Cmd+Shift+S on Mac). On Android, tap the microphone icon on the Gboard keyboard. Third-party apps like FluidVox use a global hotkey instead — hold Fn on Mac or Ctrl+Shift+Space on Windows and speak into any app.
Learn · Updated July 2026
What is voice typing? The definition, how it works, and how to turn it on everywhere
Voice typing is software that turns your speech into text in real time, directly in the app you're using. This guide covers the full picture: what voice typing means, how the technology works under the hood, how to switch it on (and off) in Google Docs, Microsoft Word, Windows 11, macOS, Android, and iPhone, which tools are free, and which ones work without internet.
Start 14-day free trialDefinition
Voice typing, defined
Voice typing software lets you speak naturally and have text appear in whatever app has focus — Slack, Gmail, Notion, VS Code, ChatGPT, Google Docs, or anything else with a text field. You talk; it types.
The category overlaps with what people used to call dictation. The difference is scope: traditional dictation tools wrote into a single dedicated app or text field (think legal dictation software, or the Dictate button inside Microsoft Word). Modern voice typing works anywhere you can type with a keyboard — which is why some tools describe themselves as a "voice keyboard." We break the terminology down further in voice typing vs dictation and define the jargon (WER, on-device, VAD, push-to-talk) in our glossary.
Concrete examples of voice typing:
- Tapping the microphone on your phone's keyboard to dictate a text message
- Using Tools → Voice typing in Google Docs to draft an essay by speaking
- Pressing Win+H on Windows 11 and dictating an email into Outlook
- Holding a hotkey in FluidVox and speaking a Slack reply, a code comment, or an AI prompt into any app
How it works
The pipeline behind voice typing
Modern voice typing tools follow a four-stage pipeline:
- Audio capture. A microphone records your speech and segments it into chunks.
- Speech recognition. A neural speech model (often from the Whisper family, or a streaming engine like Deepgram) converts audio chunks into a raw text transcript.
- AI cleanup. A language model removes filler words ("um", "uh", "like"), fixes grammar, adds punctuation, and — in the best tools — matches the tone to the app you're writing in.
- Text injection. The cleaned text is inserted into the focused text field via the operating system's accessibility APIs.
The whole pipeline runs in around a second on most modern hardware. The AI cleanup step is what separates the 2026 generation from older tools: built-in dictation like Apple's or Google's largely skips it, which is why their output reads like a raw transcript where you have to say "comma" and "period" out loud. For the deeper technical version, read how AI dictation works.
How-to
How to turn on voice typing on every platform
1
Windows 11 — press Win+H
Click into any text field and press Win+H. A small voice typing panel appears; click the microphone and speak. It's free and built in, but it processes speech in Microsoft's cloud, so it needs an internet connection. See our Windows dictation guide for its limits.
2
Mac — enable Dictation, then press the shortcut
Open System Settings → Keyboard → Dictation and switch it on. Then press the shortcut key (the microphone/F5 key, or a key you choose) in any text field and speak. On Apple Silicon Macs, Dictation can process speech on-device. Full breakdown in our Apple Dictation guide.
3
Google Docs — Tools → Voice typing
In a Google Doc, open Tools → Voice typing or press Ctrl+Shift+S (Cmd+Shift+S on Mac), then click the microphone box that appears. As of 2026 it works in current versions of Chrome, Edge, and Safari — no longer Chrome-only. Punctuation must be spoken aloud ("comma", "new line").
4
Microsoft Word — Home → Dictate
Click the Dictate microphone button on the Home tab, or press Alt+` on Windows. Dictate requires a Microsoft 365 subscription and an internet connection, since recognition runs in Microsoft's cloud.
5
Android & iPhone — the keyboard microphone
On Android, tap the microphone icon on Gboard; download offline language packs in Gboard's settings if you want it to work without internet. On iPhone, tap the microphone on the system keyboard, or install the FluidVox iPhone app and keyboard to dictate polished text into any iOS app.
6
Any app, any platform — a system-wide hotkey
Third-party tools skip per-app setup. With FluidVox you hold Fn on Mac or Ctrl+Shift+Space on Windows, speak, and cleaned-up text lands at your cursor in whatever app is focused — Docs, Word, Slack, or your IDE. Hands-free toggle modes (Fn+Space / Ctrl+Shift+H) handle longer dictation.
Turning voice typing off is generally the same control in reverse: press Win+H again or Escape on Windows, press the Dictation shortcut again on Mac, click the microphone in Google Docs so it stops listening, or tap the Gboard microphone again on Android. To remove it entirely, disable Dictation in System Settings (Mac) or voice input in your keyboard's settings (Android/iPhone).
Side-by-side
Voice typing tools at a glance (2026)
FluidVox
|
Windows Voice Typing (Win+H) | macOS Dictation | Google Docs Voice Typing | Gboard (Android) | |
|---|---|---|---|---|---|
| Price | $2.99/mo or $39 one-time (14-day free trial) | Free (built in) | Free (built in) | Free | Free |
| Works in any app | Yes — system-wide hotkey | Yes (any text field) | Yes (any text field) | No — Google Docs/Slides only | Yes (any Android app) |
| Works offline | Yes — Local plan is fully on-device | No — cloud only | Yes on Apple Silicon (on-device) | No — browser cloud service | Yes, with downloaded language packs |
| AI cleanup (filler words, grammar, tone) | Yes — 6 styles, matched per app | Auto-punctuation only | Basic auto-punctuation | No — spoken punctuation | Basic auto-punctuation |
| Custom dictionary | Yes, with auto-learned corrections | No | Limited (text replacements) | No | Personal dictionary |
| Platforms | Mac, Windows, iPhone | Windows 10/11 | macOS | Chrome, Edge, Safari (browser) | Android |
Honest take: if you dictate a few sentences a week, the free built-ins are fine — start there. Paid tools earn their price when you dictate daily and want text that arrives already cleaned up, styled for the app you're in, and with your jargon spelled correctly.
Cost
Is voice typing free?
The basic capability is free on every major platform: Windows 11 voice typing, macOS Dictation, Google Docs voice typing, Gboard, and iPhone keyboard dictation all cost nothing. For occasional use, they're genuinely good — accuracy from the big cloud engines is strong on clear speech.
Paid apps exist because the free tools stop at raw transcription. They don't remove filler words, don't adapt tone between a Slack message and a formal email, don't learn your product names, and (on Windows and in Google Docs) don't work offline. Third-party pricing in 2026 runs from FluidVox's $39 one-time Local plan to roughly $249 lifetime for Superwhisper or $8–15/month subscriptions like Wispr Flow — see our Mac roundup and Windows roundup for the full field.
Connectivity & privacy
Do you need internet for voice typing?
It depends on where the speech recognition runs:
- Cloud tools — Windows Voice Typing (Win+H), Google Docs voice typing, and Word's Dictate all send audio to a server, so they require internet and you're trusting the provider with your voice.
- On-device tools — macOS Dictation on Apple Silicon, Gboard with downloaded language packs, and FluidVox's Local plan process audio entirely on your hardware. They work offline, and audio never leaves the device.
Some apps offer both: FluidVox runs Whisper and Parakeet models locally on the Local plan and adds a cloud engine on Pro plans for those who want it. If you dictate confidential material — legal, medical, unreleased product plans — on-device is the safer default.
What modern voice typing can do
Capabilities the 2026 generation supports
- Per-app tone matching. Casual in Slack, professional in Outlook, technical in VS Code — applied automatically. FluidVox ships 6 styles (natural, casual, professional, code, notes, email) with per-app matching.
- Custom dictionaries. Add product names, jargon, and acronyms once; they're spelled correctly every time. FluidVox also auto-learns from your corrections, with AI validation.
- Multi-language support. Modern models cover 90–100+ languages; FluidVox supports 99 on-device.
- File transcription. Drop in an audio or video file and get a clean transcript out — free and on-device in FluidVox on every plan.
- Voice commands and AI assistance. Beyond typing, agents layered on dictation can rephrase, translate, or act on what you said.
- Hands-free toggles. Sustained dictation without holding a key — essential for accessibility and long-form drafting.
How close does it get to perfect? Realistically, modern engines reach word error rates around 5% on clean audio, degrading with noise, accents, and jargon — we unpack the numbers in voice typing accuracy explained.
Common use cases
Who uses voice typing
Speaking averages roughly three times faster than typing for most people, which is why voice typing has moved from an accessibility feature to a mainstream productivity tool:
- Developers — AI prompts, code comments, PR descriptions. Read more →
- Writers — first drafts, interview transcripts, long-form composition. Read more →
- Executives — high-volume email and Slack triage. Read more →
- Students — lecture transcription and essay drafting. Read more →
- Accessibility users — RSI accommodation, motor impairments, dyslexia. Read more →
How to choose
Picking the right voice typing tool
The fit depends on what you need:
- Free and occasional? Use macOS built-in Dictation, Windows Voice Typing (Win+H), Google Docs voice typing, or Gboard. They're built in and cost nothing.
- Daily dictation with AI cleanup, per-app tone, and a custom dictionary? Compare Wispr Flow, Superwhisper, and FluidVox — FluidVox is the budget pick at $2.99/month or $39 one-time.
- Mostly transcribing recorded audio (lectures, meetings, podcasts)? Look at Aiko, MacWhisper, or FluidVox's free on-device file-transcription mode.
- Privacy-critical work? Pick an on-device tool where audio never leaves your machine — FluidVox Local, macOS Dictation on Apple Silicon, or offline Gboard.
- On Linux or Android? FluidVox doesn't run there (no Linux or Android app). Gboard covers Android well; Linux users typically use open-source whisper.cpp-based tools.
Our 2026 best-of guide ranks the major options on accuracy, price, platform, and feature depth.
Voice typing FAQ
The built-in options are free: Windows 11 voice typing (Win+H), macOS Dictation, Google Docs voice typing, Gboard on Android, and iPhone keyboard dictation cost nothing. Paid third-party apps like FluidVox (from $2.99/month or $39 one-time), Wispr Flow, and Superwhisper add AI cleanup, per-app tone matching, custom dictionaries, and system-wide hotkeys that the free tools lack.
It depends on the tool. Windows 11 voice typing (Win+H), Google Docs voice typing, and Microsoft Word Dictate all process speech in the cloud, so they need an internet connection. macOS Dictation on Apple Silicon can process speech on-device, and Gboard supports offline voice typing if you download language packs. FluidVox's Local plan runs Whisper and Parakeet models entirely on your device, so it works fully offline.
Usually the same control that starts it stops it. On Windows, press Win+H again or click the microphone in the voice typing panel. On a Mac, press the Dictation shortcut key again or press Escape. In Google Docs, click the microphone icon so it stops listening, or close the Tools → Voice typing panel. On Android, tap the microphone icon on Gboard again; to remove it entirely, disable voice input in Gboard's settings under Voice typing.
On Android, voice typing is built into Gboard, Google's default keyboard. Tap the microphone icon on the keyboard, speak, and text appears in whatever app you are using. On Pixel phones and newer devices, Gboard can process speech on-device, which makes it faster and lets it work offline once you download language packs. FluidVox does not make an Android app — Gboard is genuinely the best default there.
In Microsoft Word, go to the Home tab and click the Dictate button (a microphone icon), or press Alt+` (backtick) on Windows. Dictate requires a Microsoft 365 subscription and an internet connection, because the speech recognition runs in Microsoft's cloud. If you want dictation in Word without a 365 subscription, you can use the system-level tools (Win+H on Windows, Dictation on Mac) or a third-party app like FluidVox that types into any app, Word included.
Everyday examples include dictating a text message with the Gboard or iPhone keyboard microphone, writing an essay in Google Docs with Tools → Voice typing, drafting a Word document with the Dictate button, replying to email by speaking instead of typing, and dictating Slack messages, code comments, or AI prompts with a system-wide tool like FluidVox. Anywhere you would normally type, voice typing can substitute.
Google voice typing refers to two related things: the voice typing feature in Google Docs and Slides (Tools → Voice typing, available in current versions of Chrome, Edge, and Safari) and the Gboard voice input on Android. Both are free and use Google's speech recognition. Their main limits are formatting — punctuation must be spoken aloud — and scope: Docs voice typing only works inside Google Docs, not in other apps.
They overlap heavily. Historically, dictation referred to converting audio to text in a single dedicated app or field (e.g., legal dictation). Voice typing is the modern term for keyboard-replacement dictation that works in any app. Most current tools do both.
Modern speech recognition models achieve word error rates around 5% on clean audio in supported languages. Accuracy drops with accents, ambient noise, and technical vocabulary — which is why custom dictionaries matter. Tools with AI cleanup also fix punctuation and filler words, so the text you get often needs no editing at all.
For many users with RSI, motor impairments, or accessibility needs, yes. Hands-free toggle modes and configurable hotkeys let you dictate sustained content without keyboard input. Most people land on a hybrid: voice for composing sentences and paragraphs, keyboard for editing, shortcuts, and precise cursor work.
Very likely. Most modern tools support 90+ languages — FluidVox covers 99 on-device — though quality varies, and major languages (English, Spanish, Mandarin, French, German) tend to have the best accuracy. Offline modes usually support fewer languages than cloud modes, so check the offline list if that matters to you.
That Gboard error usually means the keyboard or the Google app has been denied microphone access, or voice input has been disabled by a device policy. First check Settings → Apps → Gboard → Permissions and allow the microphone, then make sure voice typing is switched on inside Gboard's own settings. On work phones, an IT-managed profile can block voice input entirely — in that case only your administrator can enable it.
Built-in tools from Apple, Microsoft, and Google are free. Third-party voice typing apps range from $39 lifetime (FluidVox Local) to around $249 lifetime (Superwhisper) or $8–15/month subscriptions. The paid tools earn their price with AI cleanup, per-app styles, and custom vocabulary — but if you dictate rarely, start with the free built-ins.
Try FluidVox free for 14 days
Voice typing with AI cleanup, in any app, on Mac, Windows, and iPhone. Full access, no credit card required. Then $2.99/month or $39 one-time.
Start free trial