Text to Speech

Paste an article, hit Play, and this Text to Speech tool reads it out loud while it lights up each word in time with the voice, so your eyes follow along instead of drifting. That word-by-word highlight is what almost every other free reader leaves out, and it's the one reason to keep this tab open instead of clicking to the next one. It's free, no sign-up, and nothing you paste leaves your device.

Click any word in the reading view and playback jumps straight to that spot. Slow the voice down for a tricky sentence, speed it up for a long piece, or swap the voice entirely. It all runs on the voices already built into your browser and operating system, which is why it starts instantly and never uploads a thing.

1. Add your text. Type it, paste it, drag a .txt file onto the box, or click Load a text file. The counter shows how many characters you have.
2. Pick a voice. The dropdown fills with whatever voices your device offers, usually a mix of languages and accents. Match the voice language to your text, or the pronunciation goes sideways fast.
3. Set the speed from 0.5× up to 2×. About 0.9× is good for careful reading. Around 1.5× clears a backlog once your ear adjusts.
4. Nudge pitch and volume if you want. Volume here is separate from your system volume, so you can duck the speech under your music without touching either.
5. Press Play. Each word highlights as it's spoken and the view scrolls to keep pace. Pause freezes it mid-read, Resume picks up where you left off, Stop resets.
6. Jump around. Click a word in the reading view to start from there. Good for replaying one sentence you didn't catch, or skipping a long intro without deleting anything.
7. Changed a setting mid-read? Press Stop, then Play. A new voice, speed, or pitch only takes effect on a fresh start.

Most browser readers throw audio at you and leave your eyes to wander off, which is exactly what the word highlight fixes. As the voice speaks, the current word gets a solid highlight and the sentence around it a softer one, so you always know exactly where you are. Lose focus for a second and you can find your place again without rewinding.

The most obvious use is proofreading by ear. Your brain skims text you wrote and silently patches typos that aren't there, but hearing it read aloud catches the mistakes your eyes skip over. Read a draft back at about 0.9× while you watch the highlight and the doubled word, the missing "to", the sentence that runs on too long, they all surface.

It pays off for language learners too. Paste a sentence in the language you're studying, choose a matching voice, and you see precisely which word is being pronounced as you hear it. Sound and spelling snap together, which is most of the fight early on. Click that word to hear just that part again, as many times as you like.

And for anyone reading with dyslexia, ADHD, or plain tired eyes, following one highlighted word is far gentler than tracking a full line of flat text, and hearing the words at the same moment takes the load off decoding. It isn't a full screen reader, but for a single article it does the job with zero setup.

One honest caveat is worth flagging. The highlight rides on cues the voice sends about which word it's speaking, and not every voice bothers to send them. The offline voices built into Windows and macOS track cleanly in Chrome, Edge, and Safari. Some online voices report nothing, so the audio plays fine but the highlight sits still. If following along matters, pick a local voice, the ones without the "online" tag.

The tool ships no voices of its own. It asks your browser what it has and lists exactly those, and that list comes from your operating system plus your browser, so every setup differs.

• Windows includes a small built-in set (David, Zira, Mark). Install language packs in Settings and more appear on their own.
• macOS has its own family (Samantha, Alex, Daniel, and dozens more under System Settings, Accessibility, Spoken Content). The "enhanced" ones are big downloads that sound a lot better.
• Chrome and Edge layer their own Google and Microsoft network voices on top. Usually the smoothest, but they stream from the internet rather than running on your machine.
• Android and iOS each bring a different pool again.

So if a coworker swears by a voice you can't find, you probably just don't have it installed. Add it at the operating-system level, reload the page, and it shows up.

Anything tagged (online) streams its audio from a server and needs a connection, and those often sound the best of what you've got. Anything untagged runs locally on your machine, which works on a plane or in a dead zone with no lag, and they keep everything on your device. The tool never uploads your text either way, but an online voice does send the words you asked to be spoken to its provider to generate the sound. Want a voice that never cuts out when your internet flakes, or that keeps every word on your machine? Stick to local voices, which also happen to be the ones the follow-along highlight tracks best.

People often turn up wanting an MP3, a file to drop into a video or send to a friend. The browser's speech feature plays straight to your speakers and gives the page no way to save that audio, so there's genuinely no download button for these built-in voices, and a lot of sites stay quiet about it. Any browser-only tool promising a one-click MP3 of your system voices is quietly running a different cloud engine.

Here is what actually works. Record the playback with your operating system's own recorder (QuickTime on macOS, Voice Recorder or Game Bar on Windows) or Audacity with system-audio capture, then press Play here. It is crude but free, and it gets you the recording. If you need broadcast-grade files at scale, a paid cloud service will hand back an MP3 in exchange for uploading your script and giving up the fully-private part. For listening, proofreading, learning, and accessibility, the browser voices are perfect. For shipping a file, you want a recorder or a cloud service, and we say that plainly because a one-click download button for these voices would not actually work.