Speech to Text in Chrome: Built-in Options and Extensions

Chrome does not have a dictation button for the text boxes on web pages. Speech to text in Chrome comes from one of three places: your computer's own dictation, which types into whatever text box has the cursor; a website that builds in its own microphone button, such as Google Docs; or an extension you install. Each route sends your voice somewhere different, so check the route before you dictate anything private.

The built-in options are the simplest place to start, and the easiest to check. Extensions give you dictation on any page, but each one decides where your audio and text go, and you have to read its listing to find out. The rest of this guide compares the three routes, explains the permissions behind each one, and covers what to do when dictation fails.

Chrome itself has no dictation button, so speech to text comes from three places

Here is how the three routes compare. The operating system route works in any text box, the website route works only inside that site, and the extension route depends on the extension.

If you only need dictation in Google Docs or in one other web app, the built-in option is usually enough. If you need it in every app and on every site, you need the operating system's dictation, and an extension is only a fallback for pages where it is missing.

The operating system's dictation types into any text box that has the cursor

On Windows, press the Windows key and H together on a hardware keyboard to start voice typing ¹. Microsoft's requirements are a working microphone, an internet connection and the cursor in a text box ². The feature uses online speech recognition powered by Azure Speech services ³, so it does not run without a connection. Because the requirement is a text box with the cursor in it, the shortcut should work in ordinary Chrome forms. We did not test every kind of field, so try it in the one you need. For the Windows steps, see Windows voice typing.

Windows has one conflict to know about. If Voice Access is running, pressing Windows key + H does not start voice typing ⁴. Close Voice Access before you use Win+H.

On a Mac, Apple says you can speak to enter text anywhere you can type it ⁵, although some apps and secure text fields may not accept dictated text. In macOS Tahoe 26, you start Dictation by pressing the Microphone key, if your keyboard has one, or by using the Dictation keyboard shortcut ⁶. Keyboard settings show whether general text dictation, such as messages and notes (not dictation into a search box), is processed on the device and not sent to Siri servers ⁷. The on-device list covers 48 language and regional variants, and each one requires a download of speech models ⁸. Check that setting before you dictate sensitive text. For the Mac steps, see how to use Dictation on a Mac.

On a Chromebook, you turn on Dictation on ChromeOS in Settings, Accessibility, under Keyboard and text input, then start it by selecting Speak or pressing Search + d (or Launcher + d) ⁹. The language you speak has to match the device language, or dictation does not work ¹⁰. For the full steps, see our guide to Chromebook dictation.

Some websites build in their own microphone button, and Google Docs is the best known

In Google Docs, you start voice typing in Google Docs from the Tools menu by choosing Voice typing, then clicking the microphone box that appears ¹¹. Google's keyboard shortcut list gives Ctrl + Shift + S on Windows and Command + Shift + S on a Mac for starting it, and labels the shortcut as available in Chrome browsers ¹².

Google says voice typing in Docs works with the latest versions of Chrome, Edge and Safari ¹³. It also says the web browser you use controls the speech-to-text service behind the button. The support has widened over time. A December 2020 archived copy of Google's help page said voice typing was only available in Chrome ¹⁴. If you use Docs, read our guide to Google Docs voice typing for its language and accuracy limits.

Other websites can add a microphone button of their own. Those buttons are part of the site, so the site decides where the audio goes. Read the site's privacy policy before you dictate personal information into one.

Extensions add dictation to any page, but they need access to the pages you use

An extension that types into a page has to run code inside that page. Chrome calls these files content scripts and describes them as files that run in the context of web pages ¹⁵. That is what makes a single extension work across many sites, and it is also why an extension's permissions matter.

Some extensions ask for less. The activeTab permission gives an extension temporary access to the current tab only when you invoke it, such as by clicking its icon, and the access ends when you leave the page ¹⁶. An extension that uses only activeTab cannot dictate into a page you have not invoked it on.

We compared three extensions in the Chrome Web Store as checked on 2026-10-17. Each store listing shows the developer's own privacy declarations.

The Speech to Text listing shows 1,000 users ¹⁷. Its description says all data and voice transcription are handled locally or through encrypted channels ²⁰, yet its privacy section lists personally identifiable information and authentication information among the data it handles ²¹. Those two statements do not explain each other, so ask the developer or read its privacy policy before you use it with anything sensitive.

Speech Recognition Anywhere shows 30,000 users ¹⁹. It says it expands the Web Speech API in Chrome and Edge ²², which means the browser's own speech path applies to it, as described in the next section. Its privacy section declares personally identifiable information ²³.

Speechnotes shows 20,000 users ¹⁸. Its description says it uses Google's speech recognition engine for Chrome and does not collect or share your data ²⁴. Its privacy section says the developer has disclosed that it will not collect or use your data ²⁵.

These are the developers' own declarations, shown by the store. We did not verify them, so treat them as the starting point for your questions. For what such declarations can mean for your data, read is voice to text safe?. For a direct answer on whether any extension is the best choice, see the question on the best speech to text extension for Chrome.

Browser speech recognition sends audio to a server unless a page asks for on-device recognition

The default matters more than any single extension. Speech recognition on a web page is server-based by default and does not work offline ²⁶. Whether an extension or a website uses that path depends on how it was written, so the default is the one to assume unless the page says otherwise.

A browser must ask before it listens. The Web Speech API specification says user agents must only start speech input sessions with explicit, informed user consent ²⁷. Chrome's permission prompt for the microphone is one way that consent is given (see the section on permissions below).

The API has an option for on-device recognition, the processLocally property. A page can set it to true to require that recognition run on the device. When it is false, which is the default, the browser can choose local or remote processing ²⁸. MDN marks the property as experimental, so check its compatibility table before you rely on it ²⁹.

Chrome's 139 release notes describe the on-device work: it lets websites make sure neither audio nor transcribed speech is sent to a third-party service for processing ³⁰. A January 2025 reply in the Chromium design discussion estimated each on-device language pack at about 60 MB on Chrome for Windows, Mac and Linux ³¹. That is an estimate from before the feature shipped, so check the size of the pack on your own computer.

Can Chrome dictate offline? Partly. Chrome's default web speech path needs a server. On-device recognition is possible only when the page asks for it and the language pack is installed. Live Caption is a different tool. Google says it creates captions on the device and works offline, including for files on your hard drive ^{32 33}. Live Caption shows captions on your screen, and it does not type into a text box, so it is not a dictation tool. Our answer to whether Chrome can do speech to text offline goes through the cases in more detail.

Edge, Safari and Firefox handle speech differently from Chrome

Google Docs voice typing works in Edge and Safari as well as Chrome ¹³. Edge also appears in Dragon's web extension, which Nuance lists for Chrome and Edge in Dragon Professional 16 ³⁴. Nuance's Dragon Professional 16.0 release notes say that after its web extension for Edge or Chrome is installed, dictation still works but Dragon commands stop, a known issue that Nuance lists as fixed in 16.1 ³⁵.

WebKit's 2021 post on the Safari 14.1 release says Safari supports speech recognition on web pages, powered by the same speech engine as Siri ³⁶. The post does not say where that processing happens. Users need Siri turned on for the API to be available ³⁷. The post is from 2021, so check Apple's current Safari notes before you rely on the setup.

Firefox is the outlier. A Mozilla wiki page revised in January 2022 said Chrome, Edge, Safari and Opera supported the speech-to-text form of the Web Speech API, while Firefox did not ³⁸. We could not find a current Mozilla statement, so if you use Firefox, expect that a site's microphone button may not work and check Firefox's release notes before you assume otherwise.

Site microphone permissions are the first thing to check when it fails

When dictation in Chrome does nothing, check the site microphone permission first. Chrome on a computer asks you to choose Allow while visiting the site, Allow this time, or Never allow ³⁹. If you chose Never allow, the site cannot listen until you change that choice in Chrome.

The second check is the computer's own permission. On Windows 11, Microsoft says desktop apps cannot be turned on or off one by one for the microphone, and are controlled together with the Let desktop apps access your microphone setting ⁴⁰. Chrome is a desktop app, so when that switch is off, Chrome cannot use the microphone even if the site allows it. Microsoft's online speech setting is a separate switch: you turn it on or off under Settings, Privacy & security, Speech ⁴¹. This guide's sources do not say how that setting affects Chrome's web speech path.

A third cause is the input level. If nothing appears after you speak, Microsoft says to adjust the input level of your microphone ⁴².

The error codes help when you are reading a page that reports them. A network error means that network communication needed to finish recognition failed ⁴³. A not-allowed error means the browser blocked speech input for security, privacy or user preference reasons ⁴⁴. The not-allowed code points to a permission or policy, and the network code points to a connection. For a full set of fixes, read why dictation is not working in Chrome and our guide to dictation not working.

Pick the built-in option first, then the narrowest extension you can check

Start with the option that is already on your computer. On Windows, use Win+H in any text box. On a Mac, use the Dictation shortcut, and check the on-device setting if the text is sensitive. On a Chromebook, use Dictation in the Accessibility settings. In Google Docs, use Tools, then Voice typing. Windows voice typing and Google Docs both say where recognition runs, and on a Mac you can check the keyboard setting yourself.

Choose an extension only when you need dictation on pages where the built-in options do not reach. Prefer one that uses activeTab, which gives access only when you invoke it ¹⁶, over one that runs everywhere. Read its store listing, then its privacy policy. If it declares personal or login information, ask yourself whether you want that data sent to a developer you cannot check.

Before you dictate in Chrome, work through these four checks in order:

1. Built-in dictation: On Windows, press Windows key + H. On a Mac, use the Dictation shortcut. On a Chromebook, turn on Dictation under Accessibility. In Google Docs, use Tools, then Voice typing.
2. Site permission: When the site asks for the microphone, choose Allow or Allow this time in Chrome's prompt.
3. Computer permission: On Windows, make sure Let desktop apps access your microphone is on, so that Chrome is allowed to use the microphone.
4. Extension disclosures: Before you install an extension, read the privacy section of its listing and its developer's privacy policy.

If you want dictation that types into any app, not only Chrome, Yap, which we make, runs on-device on Mac and Windows. If you only need to try dictation in a browser, our free online dictation page works without installing anything. For the other ways to dictate, see dictation in Word, dictation into any app and the browser speech API, and for the wider picture, our overview of dictation in your apps.