Voice typing in Google Docs
Press your hotkey and speak; punctuated text arrives at the cursor in the document. It is the same hotkey in the tab next door, in your mail client and in anything else on the machine — and it does not care which browser you use.
Docs has voice typing in the Tools menu. Why use anything else?
Because of where it stops. The built-in one needs Chrome, Edge or Safari — not Firefox — works inside Google’s own documents and nowhere else on your machine, and expects you to say your punctuation — “comma”, “period”, “new paragraph”. This types into whatever window has focus, in any browser, with the punctuation applied from the way you speak.
We have not measured Google’s recogniser against ours and make no claim about which is more accurate. The difference on offer is where it works and what you have to say out loud. The browser and product list above is from Google’s own documentation, checked August 2026.
Whatever browser you use
Firefox, Safari, a Chromium fork, the browser your employer installed. Dictation happens outside the page, so the page does not get a say in it.
Punctuation, unspoken
Full stops, commas and sentence breaks are applied from how you speak. You are talking, not reading a document out loud with the marks named.
Docs, and then everything
The comment you leave on someone else’s draft, the mail about it, the spreadsheet cell next door. One hotkey covers the lot rather than one menu per product.
How it works with Google Docs
Three steps, none of them inside Docs.
Install the desktop client
A native binary for Windows, macOS or Linux, linked to your account once with a device code. There is no browser extension and nothing is granted access to your Google account.
Put the cursor in the document
A paragraph, a table cell, a comment, the title. Any browser will do — the client does not know or care which one is in front.
Press your hotkey and speak
Punctuated text arrives at the cursor. Switch to another tab mid-sentence and the rest lands there instead, because it follows the cursor rather than the page.
What changes when dictation is not part of the document
Google’s voice typing is a Docs feature, so it starts and stops with Docs. This runs outside the page, which is the whole difference: the same key works in the comment you leave on someone else’s draft, in the mail you send about it, and in the chat where you agree the change. One habit rather than one per product.
The cost of sitting outside is that nothing in Docs knows it is there. There are no voice commands, no “select that sentence”, no formatting by speech — Google’s built-in one does take those, and if editing by voice is what you want, it is the better tool. This types, and the editing stays where it was.
What it does not do
- No editing by voice. Google’s built-in voice typing takes commands like “select the last sentence” and “bold that”; this types words, and the editing stays on the keyboard.
- No Docs integration of any kind. It cannot see your document, cannot move the cursor and cannot read anything back — it sends keystrokes to the window in front.
- No offline mode. Audio is transcribed on our servers, so a browser tab open on a train with no signal is not where this works.
- Not on a phone or tablet. The client is desktop software; on mobile the WhatsApp bot is the path, and it returns text rather than typing it into Docs.
Common questions
Does it work in Firefox or Safari?+
Yes. The client types into whichever window has focus, so the browser is just a window with a text field in it. That is the main practical difference from the built-in voice typing, which needs one of the browsers Google supports, and Firefox is not one of them.
Do I have to say “comma” and “new paragraph”?+
No. Punctuation and sentence breaks are applied from the way you speak, so you talk the way you would to a colleague. There is no spoken command vocabulary at all, which is the reverse of the built-in one, where saying the marks is how you get them. Saying “new paragraph” will not make one — press return where you want the break.
Does it keep working when I switch tabs?+
It follows the cursor rather than the page, so moving to another tab or another application just means the text lands there instead. The exceptions across the board are programs that deliberately reject synthetic input — some games, some remote-desktop and virtual-desktop sessions, and some hardened enterprise clients.
Does it work in Sheets, Slides and Gmail?+
Anywhere a cursor sits: a cell, a speaker note, the body of a mail, a comment box. There is nothing per-product to install or enable, because nothing is installed into any of them.
Is anything installed into my Google account?+
No. There is no add-on, no OAuth grant and no permission on your Google account to review — it never learns this exists. The client runs on your machine and sends keystrokes to the window you are in.
The same client types into Word, Gmail and Slack. On a Linux desktop, where Docs in a browser is often the whole office suite, see the Linux client.
Last updated .