Text to Speech at a glance
- What it does
- Turn written text into spoken audio using the voices already installed on your device.
- Where it runs
- Entirely in your browser — no data is uploaded
- Works offline
- Yes, once the page has loaded
- Cost
- Free, with no account and no usage limit
- Category
- Text Utilities
How to use text to speech
- Paste the text you want read aloud.
- Choose a voice. The list comes from your own device, so it differs between Windows, macOS, Android and iOS.
- Adjust rate and pitch if the default pace is too quick or too slow.
- Press play, and pause or stop at any point.
What it is genuinely useful for
- Proofreading. Hearing your own writing read back catches repeated words, clumsy sentences and missing articles that your eye skips over. This is the single best use of the tool.
- Accessibility. Reading long articles aloud for anyone with dyslexia, low vision or eye strain.
- Language learning. Hearing pronunciation and rhythm in a language you are studying, if a voice for it is installed.
- Multitasking. Listening to a document while doing something else.
- Checking a script's pace. A presentation script read at a natural rate tells you whether it fits the slot.
Why your voice list looks different from someone else's
The voices come from your operating system, not from this page. Windows ships with a small set and lets you add more through language packs. macOS and iOS include a large selection, with high-quality voices available as separate downloads. Android's list depends on the Google or Samsung speech engine installed. Chrome on desktop also offers network voices, which need a connection.
If the list is short or empty, install additional voices through your system settings: Time & Language on Windows, Spoken Content under Accessibility on macOS, Text-to-speech output on Android.
Limitations worth knowing
No file download. The browser speech API plays audio; it does not expose it as a file. For an MP3 you need a dedicated service or a desktop application. Recording the output with system audio capture is the usual workaround.
Pronunciation. Names, technical terms and acronyms are frequently mangled. Spelling a word phonetically in the input is the practical fix.
Prosody. System voices apply generic intonation. They handle a plain paragraph well and struggle with questions, emphasis and dialogue.
Length. Very long inputs can be cut off by the browser. Break a long document into sections.
Privacy
Synthesis happens on your device using your system's own voices, so the text is not transmitted. One caveat: some browsers offer additional network-based voices, which do send text to the vendor. Voices marked as local, or the ones available when you are offline, are entirely on-device.
Using it to proofread
This is the use that repays the effort most reliably, and it works because of how reading works. When you read your own writing, you see what you intended rather than what is on the page — your brain supplies missing words and smooths over clumsy constructions. A synthetic voice does not.
What it reliably catches: repeated words across a line break, missing articles and prepositions, sentences that run out of breath, accidental repetition of the same phrase in consecutive paragraphs, and homophone errors such as "their" for "there" that a spellchecker passes.
A practical method: listen at a slightly slower rate than feels natural, follow along in the text, and stop at anything that makes you pause. The pause is usually the problem, even if you cannot immediately say why.
Text to speech and accessibility
The speech synthesis in your browser is the same technology screen readers use, and using it occasionally is a useful reminder of how a page sounds to someone who navigates by ear.
It is worth noting the difference, though. A screen reader does far more than read text aloud: it announces headings and their levels, describes links and form controls, reports table structure, and lets the user jump between landmarks. That is why semantic HTML matters — a <div> styled to look like a heading is silent to a screen reader, and a button built from a <span> cannot be reached by keyboard.
If you are building for accessibility, this tool is a rough sanity check rather than a test. Actual testing means using NVDA, JAWS or VoiceOver with the screen turned off.
Frequently asked questions
Not from the browser speech API, which plays audio without exposing a file. You would need a dedicated text-to-speech service or a desktop tool, or capture the system audio while it plays.
The list comes from your operating system. Install extra voices through your system's language or accessibility settings and they appear here after a page refresh.
Not for local voices, which is the normal case. Some browsers additionally offer network voices, which do send the text to the vendor - those are the ones that stop working offline.
System voices use general pronunciation rules and a dictionary that does not cover most proper nouns. Writing the name phonetically usually fixes it.
Nothing you enter here leaves your browser
Text to Speech does its work in JavaScript running on your own device. The page loads once, and after that there is no upload step and no server involved — which matters here because unpublished drafts, contracts and internal documents stay on your machine.
You can verify this rather than taking our word for it: load the page, disconnect from the internet, and the tool keeps working. Our privacy policy sets out what is and is not collected, and this guide explains why the distinction matters.