Text to Speech

Text to speech reads your writing aloud so you can hear it instead of reading it. This free tool uses the speech voices already built into your browser and device, lists British English voices first, and lets you set the speed, pitch and volume. Each sentence is highlighted as it is spoken, you can pause, resume or stop at any time, and clicking a sentence jumps straight to it. Long texts are split into sentence-sized chunks so playback does not cut out. There is no sign-up and no audio download, because the browser speech feature only plays sound through your speakers.

How to use the Text to speech

1. 1 Paste your textPaste or type up to 50,000 characters into the box, or press Try a sample to hear how it works.
2. 2 Choose a voicePick a voice from the list. British English voices found on your device appear first, followed by other English voices. Tick the box to see voices in every language.
3. 3 Set speed, pitch and volumeUse the sliders to slow the voice down for proofreading or speed it up for listening. Changes take effect from the next sentence.
4. 4 Press Play and follow alongThe current sentence is highlighted as it is read. Click any sentence to jump to it, or use Pause, Resume and Stop.
5. 5 Edit and listen againPress Stop to return to the text box, make your changes, then press Play to hear the new version.

How browser text to speech works

Modern browsers include a feature called the Web Speech API, and part of it, known as speech synthesis, can read text aloud using voices supplied by the browser or the operating system. This tool is a friendly front end for that feature. It splits your text into sentences, hands them to the browser one at a time with the voice and settings you chose, and highlights each sentence as it starts.

Because the voices belong to your device, there is nothing to install and no account to create. It also means the quality and choice of voices depend entirely on what your browser and operating system provide. The tool cannot add new voices, and it does not send your text to us. Some voices, however, are provided by the browser maker as an online service, and those may send the text to that company to generate the speech. They are marked "online" in the voice list, so choose a local voice if the text is confidential.

Why the voices differ between devices

If you open this page on two computers you may see two quite different voice lists. That is normal, and it is not something any website can change.

• Windows includes a small number of voices, which on UK installations normally means at least one British English voice, and more can be added in the Windows speech or language settings.
• Mac, iPhone and iPad offer many voices, including several British English ones, and extra or higher quality voices can be downloaded in the accessibility settings.
• Android voices come from the device's text to speech engine, often Google's, where additional languages and accents can be downloaded in the system settings.
• Chrome adds its own online voices on desktop, such as Google UK English Female and Male. Edge offers Microsoft's natural-sounding online voices. Firefox and Safari mostly use the operating system voices.
• Linux voice support varies widely and may be limited unless a speech engine is installed.

Why there is no MP3 download

Many people want to save the spoken text as an audio file, and some websites suggest they can do this with browser speech. In practice the browser speech API does not provide that. It plays sound directly through your device's audio output and gives web pages no access to the audio itself, so there is no reliable way for a page like this to capture it and offer it as an MP3 or WAV file.

Services that do offer downloads generate the audio on their own servers using commercial speech engines, which means uploading your text and usually paying for the service. If you need an audio file for your own use, options include the built-in speech or narration features of your operating system, recording your own voice, or a dedicated text to speech service with an appropriate licence for how you plan to use the audio. Check the terms before using generated voices in adverts, videos or anything published.

Speed, pitch, volume and long texts

Speed runs from half speed to double speed. Around 1× is natural for most voices; 0.8× to 0.9× is good for proofreading or for listeners learning English, and 1.3× to 1.6× suits people who are used to listening quickly. Pitch makes the voice higher or lower, and volume sets the loudness relative to your device volume. Not every voice supports every setting: some ignore pitch, and some online voices ignore both pitch and volume.

Long texts are split into chunks of a sentence or less, with very long sentences broken at commas or spaces. This is not just for the highlighting. Some browsers stop speaking part-way through a long piece of text, and pausing a single very long utterance is unreliable in others. Reading a sentence at a time avoids both problems and means changes to the sliders apply from the next sentence without starting again. The tool accepts up to 50,000 characters at once, which is roughly 8,000 words, or about an hour of listening at normal speed.

Proofreading by ear

Hearing your own writing read aloud is one of the most effective ways to catch mistakes. When you read silently, your brain fills in what you meant to write, so missing words, repeated words and clumsy sentences slip past. A synthetic voice reads exactly what is on the page, without guessing, so those errors stand out. It also gives you a sense of rhythm: writing that sounds natural when spoken is usually easier to read as well.

For the best results, slow the speed slightly, follow the highlighted sentence on screen, and stop whenever something sounds wrong. Listen for sentences that make you run out of breath, as they are usually too long, and for repeated words close together. Pay attention to punctuation too: a missing comma often shows up as a voice running two ideas together. This works well for emails, reports, web pages, tender responses and anything you are about to publish.

Accessibility and everyday uses

Listening instead of reading helps many people, including those with dyslexia, visual impairments, tired eyes or a preference for learning by ear. It is also handy for taking in a long article while cooking or tidying, practising the pronunciation of a speech, or checking how a voice assistant might read your content.

If you rely on speech output every day, a full screen reader such as NVDA or Narrator on Windows, VoiceOver on Apple devices or TalkBack on Android will serve you better, because it reads everything on screen, including menus and buttons. For website owners, trying your own pages with a screen reader is a valuable accessibility check, and it often reveals unlabelled buttons, missing alt text and confusing headings that a quick visual check misses.

Teachers and parents also use browser speech to help children follow along with a passage, and language learners use it to hear how a sentence should sound before trying it themselves. Changing the voice to a different accent of English can be a useful exercise in its own right.

Examples you can copy

Proofreading           0.8x - 0.9x
Normal listening       1.0x
Familiar material      1.2x - 1.4x
Experienced listeners  1.5x - 2.0x

500 words    about 3-4 minutes
1,500 words  about 10-11 minutes
8,000 words  about an hour

Common mistakes to avoid

Frequently asked questions