Text to Speech

Turn any text into natural speech with your device voices — adjust speed, pitch, and volume.

25 words10s of audio

Speech is generated by voices installed on your device, so nothing you type is sent to a server. The voices available depend on your operating system — Windows, macOS, Android, and iOS each ship a different set, and some browsers offer additional online voices marked as such.

What is Text to Speech?

A text to speech tool reads written text aloud in a synthetic voice. Paste an article, a script, a study note, or an email draft, choose a voice, and listen. This one uses the Web Speech API, which means the speech is generated by voices already installed on your device rather than by a remote service — nothing you paste is transmitted anywhere, and there is no character limit, no quota, and no signup.

The practical uses go well beyond novelty. Hearing your own writing read back is the fastest way to catch clumsy sentences that read fine on the page. People with dyslexia or visual impairments use TTS to consume long documents. Language learners use it to hear pronunciation. Commuters turn articles into something they can listen to instead of read. And anyone proofreading a script will find that the ear catches what the eye skims past.

Speed, pitch, and volume are all adjustable, and the voice list is grouped by language — typically anywhere from a handful to over a hundred voices depending on your operating system. Long text is automatically split into sentence-sized chunks before being spoken, which works around a well-known browser limitation where a single long utterance can be cut off part-way through.

Use Cases

Here are the most common ways people use Text to Speech every day.

Proofreading Your Own Writing

Reading your own draft silently is unreliable — your brain supplies the words you meant rather than the ones you typed. Hearing it spoken aloud exposes run-on sentences, repeated words, missing articles, and rhythms that do not work. Professional editors have used read-aloud passes for decades. Paste your draft, set the speed slightly below normal, and listen with the text in front of you.

Accessibility and Reading Support

For readers with dyslexia, low vision, ADHD, or visual fatigue, hearing text alongside seeing it substantially improves comprehension and endurance. This is not a replacement for a full screen reader like NVDA or VoiceOver, which navigates entire interfaces — but for reading a specific document or article, it is quicker to set up and easier to control.

Learning Pronunciation in Another Language

The voice list is grouped by language, so you can select a native voice for Spanish, French, German, Hindi, Mandarin, or whatever you are studying, and hear how a phrase should sound. Slowing the rate to 0.7× makes individual sounds easier to distinguish. Bear in mind that synthetic voices approximate natural speech rather than reproducing it, so use this alongside recordings of native speakers rather than instead of them.

Listening to Articles While Doing Something Else

Long-form articles, documentation, research papers, and reports can be listened to during a commute, a walk, or routine work. At 1.25–1.5× speed most people can follow comfortably while doing something that does not demand language processing. Paste the text, press play, and put the device down.

Rehearsing Scripts and Presentations

Hearing a script spoken at a fixed pace tells you how long it actually runs and where the phrasing fights the speaker. Voiceover artists, podcasters, and presenters use TTS as a first pass before recording — it costs nothing to hear a draft five different ways before committing to a take.

Checking Content for Screen Reader Users

If you write web copy, alt text, or documentation, hearing it spoken shows you how it lands for someone using assistive technology. Acronyms, unusual punctuation, symbol-heavy text, and long unbroken URLs all sound very different from how they look. This is a fast, rough approximation of a real screen reader test — useful, but not a substitute for testing with the real thing.

Examples

Example 1

Proofreading a Blog Draft

Paste a 1,200-word draft, set speed to 0.9×, and follow along with the text.

Input 1,200-word article draft
Output About 8 minutes of speech — repeated phrases and overlong sentences become obvious
Example 2

Hearing a Phrase in Spanish

Select a Spanish voice from the language-grouped list and slow the rate down.

Input "¿Dónde está la estación de tren?" at 0.7×
Output Slow, clearly enunciated Spanish pronunciation
Example 3

Listening to Documentation at Speed

Paste a technical page and raise the rate for a faster listen.

Input 3,000 words at 1.5×
Output About 13 minutes instead of 20

Text to Speech vs a cloud TTS service

Paid services like ElevenLabs or Amazon Polly produce better audio and downloadable files. Here is the trade-off.

Feature Toolorah a cloud TTS service
Cost Free, unlimited Per-character pricing or a subscription
Voice quality Depends on your OS voices Consistently high — neural voices
Downloadable audio file No Yes
Text privacy Stays on device with local voices Uploaded to their servers
Character limits None Quota-based
Signup required No Yes
Works offline Yes, with local voices No

Tips for Using Text to Speech

  • Voices marked "(online)" are provided by the browser vendor and usually sound better, but they require a network connection. Unmarked voices are installed locally and work offline.
  • For proofreading, set the speed slightly below 1× — errors are much easier to notice when the pace is unhurried.
  • If you want more or better voices, install additional language packs through your operating system settings. They appear here after a browser restart.
  • Punctuation drives the phrasing. Adding commas and full stops to a run-on passage improves the read noticeably.
  • Chrome loads its voice list asynchronously. If the dropdown looks short right after the page opens, wait a second and it will fill in.

Frequently Asked Questions

Is my text sent to a server?

No for locally installed voices, which is most of them — speech is generated on your device by the operating system, and the text never leaves your machine. Voices labelled "(online)" are the exception: those are synthesised by the browser vendor's servers, so text sent to them does leave your device. If you are reading anything confidential, pick a voice without the online label. Toolorah itself never receives, stores, or logs your text either way.

Why do I only have a few voices?

The voice list comes entirely from your operating system and browser, not from this site. Windows ships with a small default set and many more available through Settings → Time & Language → Speech. macOS and iOS have a broad selection under Accessibility → Spoken Content → Voices, including high-quality downloadable ones. Android varies by manufacturer. Chrome typically exposes the most voices, Firefox the fewest. Installing a language pack and restarting your browser is the usual fix.

Can I download the speech as an MP3?

Not directly. The Web Speech API sends audio straight to your speakers and does not expose it as a recordable stream, so there is no reliable in-browser way to save it as a file. If you need an audio file, use your operating system's built-in recorder capturing system audio, or a dedicated TTS service that returns a downloadable file. This tool is built for listening rather than producing assets.

Is there a character limit?

No limit is imposed. Very long text is automatically split into chunks of roughly 200 characters at sentence boundaries before being spoken, which works around a long-standing Chrome bug where a single long utterance stops part-way through with no error. You will not notice the split — it follows natural sentence pauses.

Why does the voice sound robotic?

Quality varies enormously between voices. The older built-in voices on Windows use concatenative synthesis and sound noticeably synthetic. Newer neural voices — Microsoft's "Natural" voices, Apple's Siri and Premium voices, and Google's online voices — are dramatically better. Look for voices labelled Natural, Enhanced, Premium, or "(online)" in the list, and install them through your OS settings if they are not there.

Does it work offline?

Yes, with locally installed voices. Once the page has loaded, everything runs on your device. Voices labelled "(online)" need a connection because they are synthesised remotely. If you expect to be offline, select a local voice and test it before you need it.

Can it read in languages other than English?

Yes — the voice list is grouped by language code, and whatever languages your device has installed will appear. Common ones include Spanish, French, German, Italian, Portuguese, Hindi, Arabic, Mandarin, Japanese, and Korean. The catch is that a voice reads text using the pronunciation rules of its own language, so an English voice reading French text will sound wrong. Match the voice to the text.

Why does playback stop unexpectedly?

Usually one of three things. Switching browser tabs can suspend speech in some browsers. Changing the voice, speed, or text mid-playback deliberately stops it so the new setting takes effect from the start. And online voices will cut out if your connection drops. Press Stop then Play to restart cleanly.