1. Choose a browser Google supports on Windows and Linux
Google’s help page says voice typing works with the latest versions of Chrome, Edge and Safari. Safari is Apple’s browser, so on a Windows or Linux PC the practical choices are Chrome and Edge. Firefox is not on the list, and neither is any other browser; if voice typing is missing or the microphone does nothing, opening the same document in Chrome or Edge is the first test.
On Linux, Google Chrome is the straightforward option. Google publishes it for 64-bit Ubuntu 18.04 or later, Debian 10 or later, openSUSE 15.5 or later and Fedora Linux 39 or later, as .deb and .rpm packages for x86-64 and arm64. Google’s voice typing page names browsers rather than operating systems, so try it on your own setup. It also notes that a Google Workspace administrator can turn the feature off for an organization.
Why the browser matters: Google says that when you turn on voice typing, your web browser controls the speech-to-text service and decides how your speech is processed before sending the text to Docs. Mozilla’s developer documentation adds that in Chrome, speech recognition on a web page uses a server-based engine, so the audio goes to a web service and recognition won’t work offline.
2. Turn on voice typing with Tools or Ctrl+Shift+S
Make sure the computer’s microphone is on and working, then open the document in Docs. Choose Tools > Voice typing; a microphone box appears. Click the microphone when you are ready, speak clearly at a normal pace, and click it again when you finish. In Chrome on Windows or Linux, Ctrl+Shift+S starts voice typing without the menu; Google’s shortcut list marks it as available in Chrome browsers.
You don’t have to stop to fix a mistake. Move the cursor to the wrong word, correct it with the keyboard, then move the cursor back and keep talking. Words underlined in gray have alternatives: right-click one to see Google’s suggestions. Slides has a related command, Tools > Dictate speaker notes, but the editing commands below don’t work there.
Chrome keeps its own microphone choice, separate from the one Windows or Linux uses by default. If voice typing listens to the wrong device, open Chrome’s menu, go to Settings > Privacy and security > Site settings > Microphone, pick the default microphone from the drop-down, and make sure docs.google.com is not on the not-allowed list.
- Tools > Voice typing, then click the microphone.
- Ctrl+Shift+S in Chrome.
- Click the microphone again, or say “Stop listening”, to stop.
- Right-click gray-underlined words for alternatives.
3. Docs voice commands: English only, typing in 100+ languages
Google lists 119 language and accent options for voice typing, from Afrikaans and Amharic through many variants of Arabic, English and Spanish to Vietnamese and Zulu. Spoken punctuation, “period”, “comma”, “exclamation point”, “question mark”, “new line” and “new paragraph”, may not be available in every language.
Editing and formatting commands are English only, and Google requires both your account language and the document’s language to be English. With that in place you can say “select paragraph” or “select last 3 words”, “bold”, “apply heading 2”, “create numbered list”, “insert comment” followed by your comment, “insert table 3 rows by 4 columns”, “go to end of paragraph” or “scroll down”. “Stop listening” ends voice typing, “resume” moves the cursor to the end of the paragraph and starts again, and “voice typing help” opens the command list.
When commands come out as text, Google’s advice is to speak slowly and pause before and after each command; the command’s words can appear in the document for a moment before it runs. The microphone shows a bubble with the last command it heard, and saying “undo” reverses a wrong one.
4. Fixing voice typing that stops on Windows or Linux
For the message “We’re having trouble hearing you,” Google suggests a quieter room, an external microphone and a higher input volume. If the microphone doesn’t work at all, it suggests checking that it isn’t broken or in use by another application, checking the system’s microphone settings, and restarting the computer.
On Windows 11, those settings are under Settings > Privacy & security > Microphone. Microsoft says Microphone access and Let apps access your microphone should be on, and a browser that isn’t listed individually counts as a desktop app, controlled by Let desktop apps access your microphone. In Chrome, also check the site permission: choose Allow when Docs asks, and if you blocked it earlier, allow docs.google.com under Site settings > Microphone. On a managed work or school Chrome, the administrator may have locked these settings.
On Linux, set the right input device in your desktop’s sound settings and in Chrome’s default-microphone drop-down, and close other apps that might hold the microphone. If voice typing still fails, try a new document to separate a problem with one file from a problem with the browser.
- Wrong browser: switch from Firefox to Chrome or Edge.
- Blocked site: allow the microphone for docs.google.com.
- Silent input: check the system input device and Chrome’s default microphone.
- Missing menu at work: ask whether an administrator turned voice typing off.
5. Windows voice typing and Voice access as Docs alternatives
Windows has two system-level options that don’t depend on Docs’ own feature. Windows key+H starts voice typing in whatever field has the cursor; Microsoft says it uses online Azure Speech recognition and needs an internet connection and a cursor in a text box, and Windows 11 lists 46 languages for it. It types into Docs like a keyboard, so it doesn’t understand Docs commands such as “apply heading 2”; it has its own short list, such as “delete that” and “select that”.
Voice access, in Windows 11 version 22H2 and later, works without an internet connection and also lets you control the PC by voice. Say “Dictation mode” before dictating into a document, add unusual names with “Add to vocabulary”, and use Alt+Shift+B to wake it or put it to sleep. Microsoft documents both tools for text fields in general rather than for Google Docs, so try them in your document before you depend on them.
6. IraVoice in Google Docs on Windows and Linux, in any browser
Docs voice typing is free and is the right first choice when Chrome or Edge suits you and its processing is acceptable. IraVoice is a different kind of tool: a desktop app that recognizes speech on the computer and types into whatever has focus, so it doesn’t matter whether Docs is open in Chrome, Edge or Firefox. IraVoice 1.2.2 runs on macOS 13+ on Apple silicon, Windows 10 or 11 (64-bit), and Linux (beta) with NVIDIA Parakeet; the Parakeet Unified EN model (about 700 MB) downloads once, and recognition then works offline.
On Windows, click into the document, hold Right Ctrl, speak, and release: IraVoice pastes the words at the cursor with Ctrl+V and restores your clipboard. On Linux the default key is Ctrl+Space, and IraVoice types the words with a helper, xdotool on X11 or wtype, kwtype, dotool or ydotool on Wayland, which you may need to install. On Wayland, if the key doesn’t reach IraVoice, bind iravoice --toggle-transcription to a GNOME or KDE custom shortcut.
IraVoice types words; it does not run Docs’ formatting commands, so headings and lists stay on the keyboard. Replacements and snippets let one spoken phrase type exact text, custom words help with names, and when it mishears something, Ctrl+Alt+Shift+F opens the fix so it types that word correctly from then on. The Linux version is a beta: its AppImage runs on Fedora, openSUSE and any other 64-bit distribution from 2022 or later, and on Fedora you can install the .rpm instead.
IraVoice Local costs $79 once for one computer, $109 for two, or $149 for three, and the 3-day free trial begins with your first dictation, with no email or card needed to download.