ken107/read-aloud: a one-click text-to-speech extension for Chrome and Firefox

What Read Aloud actually solves, and for whom

The problem is narrow and real: a page of text is on screen, and you would rather hear it than read it. Read Aloud is a Chrome and Firefox extension that converts webpage text to audio, and the README lists its intended audience explicitly: people who prefer listening to reading, people with dyslexia or other learning disabilities, children learning to read, and anyone who wants an alternative way to consume web content. The README also names the kinds of sites it is built for, including news sites, blogs, fan fiction, publications, textbooks, school and class websites, online universities and course materials. That list is the honest scope statement. This is not a document reader for your filesystem, not a screen reader that navigates application interfaces, and not a service that turns an uploaded PDF into an audio file. It reads what the browser has already rendered, and the package description in package.json says it supports 40+ languages. The extension is published on the Chrome Web Store and on Mozilla Add-ons, so the normal user never touches the source tree.

How the extension turns a rendered page into speech

The repository layout shows a standard Manifest V3-era WebExtension split. background.js holds the service worker logic, page-scripts/ holds code injected into the pages being read, and the user-facing surfaces are separate HTML files: popup.html for the toolbar menu, player.html for the playback controls, options.html and advanced-options.html for settings, custom-voices.html for voice configuration, languages.html, shortcuts.html, connect-phone.html, pdf-viewer.html and report.html. The _locales/ directory carries translations, and js/ plus css/ hold the shared code and styling. Two mechanisms are visible in the README. Browser-native voices come from the platform's own speech synthesis, which is why the free path needs no account. Cloud voices come from named providers: Google Wavenet, Amazon Polly, IBM Watson and Microsoft. The README states that some cloud-based voices may require an additional in-app purchase to enable, and it links a separate document at docs/usage/premium-voices.md for Google Wavenet and Amazon Polly. The package.json scripts add one more detail about how extraction is delivered: a sync-page-scripts script copies js/messaging.js into page-scripts/ and pushes that directory to an S3 bucket, which means site-specific extraction logic is shipped separately from the extension bundle and can be updated without a store release. That is a sensible design for a tool whose hardest part is not speaking but deciding what on a page is worth speaking.

Installing it and getting the first page read

For almost everyone the install is a store install. Chrome and Chromium-based browsers get the extension from the Chrome Web Store, and Firefox gets it from Mozilla Add-ons. There is no npm install step for users, and no server to run.

mkdir build
npm run-script package

Those two commands are the Firefox source build documented in the README. The first creates a build directory, the second runs the package script from package.json, which zips _locales, css, img, js, sound, the HTML files, manifest.json and background.js into build/package.zip. Extract that zip and you should see a manifest.json.

about:debugging

Type that in the Firefox address bar to reach the debugging page. Click "This Firefox", then "Load Unpackaged Extension", then select the manifest.json you extracted. The README warns to make sure no existing Read Aloud add-on is already installed first, which is the step people skip and then wonder why nothing changed.

ALT/Option + P           : Play/Pause
ALT/Option + O           : Stop
ALT/Option + Comma       : Rewind
ALT/Option + Period      : Forward

Once installed, the README documents two ways to start: the extension button in the toolbar, or the right-click context menu. The shortcuts above control playback afterwards. To change anything, click the Read Aloud icon, stop any text that may be playing, then click the gear icon in the context menu. The README notes it may take a second or two for the settings to appear, which is worth knowing before you conclude the button is broken.

Where Read Aloud stops being the right tool

The first limitation is the one the README states most plainly: some cloud-based voices require an additional in-app purchase. The free experience is whatever your browser's speech synthesis provides, and voice quality, language coverage and availability vary by platform rather than by anything this project controls. If your reason for choosing Read Aloud is a specific premium voice, you are buying into a provider relationship, not just an extension.

The second limitation is scope. The search questions around this project are full of Word, PowerPoint, Adobe Acrobat, Microsoft Edge and iPhone, and the README addresses none of them. Read Aloud is a Chrome and Firefox extension. If your reading happens inside a desktop application, this is the wrong tool and no configuration will change that.

The third is the release cadence visible in the repository metadata. The most recent release listed is v1.54.1 from 2022-03-09, tagged as undoing the previous release, with v1.54.0 the same day described as making Google Docs work better. Before that, ff-1.45.1 from 2021-08-24 is described as a bug fix. The repository itself has not been archived and the last push was on 2026-07-10, so work on the tree continues, but a reader who equates store version numbers with active feature development should look at that gap directly. The README also does not document any rollback procedure for a bad release, which is a gap if you plan to pin a version in an enterprise deployment.

Finally, page extraction is inherently fragile. The sync-page-scripts script exists precisely because sites change their markup and extraction has to be adjusted. Expect some pages to read badly: navigation, sidebars and cookie banners are exactly the kind of text a naive extractor picks up.

Read Aloud against a browser's built-in reader

The obvious alternative is the reading feature already inside the browser. Microsoft Edge ships an immersive reader with a read-aloud mode, and the search questions show people looking for exactly that. The difference in approach is architectural. Edge's reader is part of the browser, so it gets the rendering pipeline, the OS voices and the update channel for free, but it only exists in Edge and only on pages the reader can parse. Read Aloud is a WebExtension that installs into Chrome, Chromium-based browsers and Firefox, so the same tool follows you across two engine families, and it can offer cloud voices from Google Wavenet, Amazon Polly, IBM Watson and Microsoft on top of the local ones. The trade is that you depend on a third-party extension for something the browser maker may already provide, and on a project whose newest listed release is from 2022. If you are already standardized on Edge and only need the built-in voices, the browser feature is the lower-dependency choice. If you move between Chrome and Firefox, or you want a single playback UI with keyboard shortcuts and a player window, the extension is the one that travels.

Licence, maintenance and the cost of upgrading

The licence is MIT, stated in both LICENSE and the package.json license field, with the author listed as Hai Phan. MIT is permissive: you can fork, modify and redistribute, including commercially, provided the copyright notice and permission notice are preserved. That matters here because the extension is distributed through two stores, and a fork would need its own store listing and its own review process. Nothing in the repository suggests the project offers a hosted service you would be locked into. Premium voices are billed by the third-party provider the README names, not by this repository, so the cost model sits outside the MIT grant. This is not legal advice; read LICENSE and the provider terms yourself.

The upgrade cost is the practical question. The store channels mean users get updates automatically, so the maintenance burden is close to zero for individuals. For anyone building from source, the Firefox path is the documented one, and it depends on npm run-script package plus a manual about:debugging load, which is a per-developer ritual rather than an automated pipeline. The repository does not document a signed .xpi build, a CI release workflow, or a version-pinning strategy. If you need reproducible builds with provenance, that work is yours to add. The separately synced page-scripts directory is a second moving part: extraction logic can change on the S3 side without any corresponding change to the extension version you installed, which is convenient for fixes and awkward for anyone trying to audit exactly what code ran on a given day.

Editorial conclusion

Adopt Read Aloud if you want a click-to-speak extension for Chrome, Chromium-based browsers or Firefox and you are satisfied with the voices the browser already ships, because that path needs no account and no payment. Skip it if you need a reading tool inside Microsoft Word, PowerPoint, Adobe Acrobat or the iPhone system, since this project is a browser extension and the README documents nothing outside that scope. Before you commit to premium voices, open the premium-voices document in docs/usage and check which providers are listed, then confirm in the extension's settings that the provider you want is actually offered there. Verify first that the packaged build installs cleanly in your browser version, because the newest release listed on the repository is v1.54.1 from 2022-03-09.

Frequently asked questions

Official sources

1. ken107/read-aloud on GitHub
2. License: MIT
3. Project website
4. README
5. Releases