Blog · Tools
Dyslexia-friendly text-to-speech in Chrome
Text-to-speech is the single most underused tool in the dyslexic reader's toolkit. Most people who try it abandon it inside ten minutes - the default voice sounds robotic, the controls are clunky, and reading by ear feels slow compared to skimming by eye. None of those are real problems. They are setup problems, and they all have fixes. Here is the TTS setup that actually works in Chrome, the built-in option most people never find, and the combination that beats either reading or listening on its own.
The short answer
For most dyslexic readers, the best Chrome setup is the built-in Read aloud feature in Reading mode, paired with a dyslexia-friendly font applied through an extension like LexiFont. It is free, requires no account, runs offline once the voice is downloaded, and supports highlight-as-it-reads which is the feature that actually moves the needle for comprehension.
If you need something more powerful - PDFs, longer queues, premium voices - NaturalReader, Speechify and Read&Write are the three commercial extensions worth your time. Stop trying every extension on the Chrome Web Store. Most of them are thin wrappers around the same system voice you already have.
Why TTS helps dyslexic readers (and where it does not)
Dyslexia is, at heart, a phonological-processing condition - the brain has trouble mapping written letters to sounds quickly and accurately. Reading aloud bypasses the slowest part of that pipeline. You still have to follow the text with your eyes, but the decoding work is offloaded onto the synthesised voice, which frees up working memory for comprehension. See our piece on dyslexia and working memory for why this matters more than it sounds.
The catch: TTS only helps if you follow along visually while listening. Pure audio - eyes closed, just the voice - is fine for podcasts but is a worse way to read than reading itself. Without the visual anchor, you lose your place, miss structure (headings, lists, emphasis), and cannot easily re-read a sentence. The mode that works is bimodal reading: eyes on the text, ears on the voice, ideally with the current word highlighted as the voice reads it. That is the configuration we will set up below.
What TTS does not fix: layout confusion, columns, footnotes, equations, image-only PDFs, and most things where the page itself is the problem. For those, see our separate guides on PDF reading in Chrome and reading research papers.
The built-in option most people miss
Chrome shipped a Read aloud feature into its Reading mode in 2024 and most people still do not know it exists. It is the right starting point because it is free, requires no extension, and supports highlight-as-it-reads on any web page that Reading mode can extract.
To turn it on:
- Open a long article in Chrome. Click the menu (three dots, top right) and choose Reading mode. The page opens in a clean side panel.
- At the top of the Reading mode panel, click the play button. The browser starts reading aloud, and the current sentence (and within it, the current word) is highlighted as the voice progresses.
- Click the gear icon in the panel to change the voice, the speed, and - critically - the highlight colour. The "highlight on" setting is what makes this work.
The default voice on Windows and Mac is the system voice, which is fine but unremarkable. Chrome on a recent install also offers downloadable "natural" voices that sound noticeably more human. Open chrome://settings/accessibility and look for Text-to-Speech voices - the natural voices add 50-100 MB per language but they are worth it. Robotic voices are the number one reason people abandon TTS, and they are not the only option.
Reading mode + Read aloud handles maybe 80% of long-form web reading: news articles, blog posts, Wikipedia, documentation. It does not handle PDFs in the browser well, it cannot queue more than the current article, and it is not available on every site (highly interactive pages defeat the extractor). For those gaps you need an extension.
The three TTS extensions worth installing
1. NaturalReader - the workhorse
NaturalReader is the closest thing the Chrome ecosystem has to a default TTS extension. Free tier gives you the standard system voices and unlimited basic reading. The paid tier ($9.99-$19.99/month depending on plan) unlocks premium AI voices that are noticeably better - the kind you would not be embarrassed to listen to for an hour. It reads selected text, full pages, and PDFs loaded in Chrome's PDF viewer, with word-by-word highlighting.
The strength of NaturalReader is breadth. It works almost everywhere - Gmail, Google Docs, Wikipedia, news sites, PDF viewer, even text inside images on the paid OCR tier. The weakness is that the free version's voices are no better than Chrome's built-in ones, so you are mostly paying for the smoother integration and the premium voices.
Recommended if: you want one extension that handles everything, and you have read enough on the free tier to know whether the premium voices are worth the subscription.
2. Speechify - the polished option
Speechify is the most consumer-friendly of the three. Its interface is clean, its premium voices include reasonable celebrity-trained options (Snoop Dogg, Gwyneth Paltrow, etc. - novelty, but the regular AI voices are also excellent), and it syncs across devices including a mobile app. Pricing is steeper - typically $11.58/month billed annually for the premium tier, which is the only tier that includes the better voices.
Speechify's strongest feature for dyslexic readers is its reading speed controls. You can train yourself up from 1x to 1.5x or 2x over weeks, and many users end up reading-by-listening faster than they read visually. The free tier limits you to about 10 minutes of premium voices per day, which is enough to evaluate but not enough to use it as a daily driver.
Recommended if: you want TTS that feels like a polished consumer product and you are willing to pay for it.
3. Read&Write by Texthelp - the accommodations tool
Read&Write is the extension institutions buy. If you are a student or employee with documented dyslexia, your school or workplace may already license it - check first before paying. It includes TTS but also a screening pen, vocabulary tools, prediction, and a study-skills layer that the consumer extensions do not match. The interface is dense compared to NaturalReader and Speechify, and the standalone consumer subscription (around $145/year) is the steepest of the three.
Recommended if: you need formal accommodations integration, you are studying, or your institution provides the license.
The combo that beats either tool alone
Here is the actual recommendation, the one that improves outcomes the most. Use TTS and a dyslexia-friendly font, together, on the same page. Listen and read at the same time.
The reason this works so well is that decoding load and comprehension load are competing for the same finite working memory. TTS removes most of the decoding load. A dyslexia-friendly font removes the rest. What you have left is pure comprehension, which is the thing you were trying to do in the first place. See our 2026 dyslexia font guide for which face to use.
Setting this up takes about three minutes:
- Install LexiFont and pick a dyslexia font - OpenDyslexic if letter rotation is your issue, Lexend or Atkinson Hyperlegible if you mainly want clearer letterforms without the weighted look.
- Pick your TTS - either the built-in Chrome Read aloud or one of the three extensions above. Turn on word-level highlighting.
- Open a long article. Apply the font with LexiFont. Start the TTS. Read along at 1.0x for the first paragraph, then nudge the speed up by 0.1x at a time until you find the pace where you can still follow without straining.
Most readers find their comfortable bimodal speed lands between 1.1x and 1.4x. Beyond 1.5x the voice starts compressing diphthongs and the highlight can no longer keep up with where your eye is. Below 1.0x the voice feels patronising and your eye outruns it constantly.
Voice choice matters more than you think
The default Microsoft and Apple system voices have improved a lot in the past three years but they are still recognisably robotic on long readings. Robotic voices fatigue dyslexic listeners faster than human-sounding ones - your brain works slightly harder to map the prosody (stress, rhythm, intonation) to meaning, and that work compounds over a 30-minute article.
If you can only change one thing about your TTS setup, change the voice. The current best free options on each platform:
- Windows 11: the Natural voices (Ava, Andrew, Aria) downloaded via Settings → Accessibility → Narrator → Add natural voices. These also work in Chrome's Read aloud.
- macOS: the "Premium" and "Enhanced" voices under System Settings → Accessibility → Spoken Content. Ava, Tom, and Samantha at the Premium tier are the standouts.
- Chrome OS: the Google natural voices ship with the system and are good out of the box.
For premium AI voices you need an extension subscription - the ElevenLabs-class voices that NaturalReader and Speechify license are a clear step up from anything that runs locally. They are not necessary, but they are nice.
Speed - the unintuitive part
New TTS users almost always pick a speed that is too slow. The instinct is to set the voice to 0.9x or 1.0x because "I want to follow along carefully," but at that speed your eye is constantly waiting for the voice and you spend cognitive effort on patience instead of reading.
The right calibration: pick a speed where you have to slightly work to keep up. Not strain - just gentle work. That speed is usually 1.2-1.4x for new bimodal readers and creeps up to 1.5-2.0x for experienced users over a few months. The faster you go, the more the dyslexia font matters, because at speed your eye has less time to recover from a letter confusion.
One more thing: never let the speed creep up past the point where you have stopped following with your eye. The moment you become an audio-only listener, you have lost the comprehension benefit. If you find yourself zoning out, slow down by 0.2x and re-anchor your eye on the highlighted word.
Common TTS problems and the fixes
"The voice reads punctuation as words." Switch voices - the older Microsoft and Apple voices do this; the natural ones almost never do. If you are stuck with an older voice, NaturalReader and Speechify both have a "skip punctuation" toggle.
"It reads the menu and the cookie banner." Use Chrome's Reading mode first to extract the article, then start TTS inside the cleaned view. The Read aloud built into Reading mode does this automatically. For extensions, most have a "selected text only" mode - select the article body before pressing play.
"It cannot read this PDF." Most likely the PDF is a scanned image, not text. Run it through OCR first - our PDF reading guide covers the recipe. NaturalReader Pro and Speechify Premium both include OCR but the free options will not see the text.
"It pronounces names and technical terms wrong." Mostly unfixable in TTS without manual pronunciation dictionaries (Read&Write has these; the consumer extensions mostly do not). For research-heavy reading where mispronounced jargon breaks the flow, mute the TTS for those sections and read them visually.
"It loses my place when I switch tabs." Known limitation - most extensions stop when their tab is backgrounded. Speechify keeps reading across tab switches on its paid tier; the free options do not. The Reading mode panel persists for the article it was opened on, but switching to a different article resets it.
Privacy - what gets sent where
Chrome's built-in Read aloud uses local voices once you have downloaded them - nothing leaves your machine. NaturalReader and Speechify send selected text to their cloud for the premium AI voices (that is how those voices stay so good); their privacy policies cover this. Read&Write's enterprise deployment is usually self-hosted or on a regional cloud the institution controls.
If you are reading something confidential - legal documents, medical records, work-internal material - prefer the local voices. Chrome's Read aloud with natural voices is the safest, fastest answer.
What about TTS on mobile?
Mobile is a different setup and the OS-level features beat any browser extension. On iOS, turn on Speak Selection and Speak Screen under Settings → Accessibility → Spoken Content. On Android, use Select to Speak under Settings → Accessibility. Both highlight as they read and work in any app, not just the browser. See our mobile reading guide for the full setup.
The case for trying TTS again if you have given up
A lot of dyslexic adults tried TTS in school, hated the voice, and have not touched it since. The voices in 2026 are not the voices of 2016. The natural and premium voices on every major platform are now close enough to human that the artificial-prosody fatigue is mostly gone. If your last attempt was more than three years ago, the experience has materially changed.
Try this concrete test: open a long article you have been putting off (the kind of piece you keep saving to read later and never do). Apply your dyslexia font with LexiFont. Start Chrome's Reading mode and press Read aloud with a natural voice at 1.2x. Read along visually. If you reach the end of an article you would otherwise have abandoned, you have your answer.
Get LexiFont Pro - OpenDyslexic, Lexend, Atkinson Hyperlegible and Comic Neue for $14.99 one-time