Text to Speech — Read Any Text Aloud in Your Browser
Paste text, pick a voice and speed, press play. Sentences light up as they are read. No account, no upload.
🔒 Runs in your browser — files never uploaded ⚡ No signup 💯 Free
Loading voices…
Speech is generated by your device's own voices and played through the speakers. The browser API offers no audio file, so there is nothing to download — to keep a recording, use your system's audio or screen recorder while it plays.
Have any text read aloud — an article, a chapter, an email draft, notes for revision — using the voices already built into your device. Choose from every voice your system offers, in dozens of languages, set the speed and pitch, and follow along as each sentence is highlighted. Pause and resume at any point.
The text never leaves your browser; speech is generated locally by the Web Speech API. That also means there is no audio file to download — the page is a reader, not a recorder.
How to use Text to Speech Reader
- Paste the text. The word count and an estimated reading time appear underneath.
- Choose a voice. They are grouped by language; the list is whatever your operating system and browser provide, so it differs between a Mac, a Windows PC and a phone.
- Set the speed — 1.0 is normal speech, 1.3–1.5 suits proof-listening, 0.8 suits language learners.
- Press Play. Pause and Stop do what they say; the current sentence is highlighted as it is spoken.
Where the voices come from
Browsers expose the operating system's speech engines through the Web Speech API. On Windows those are the Microsoft voices (Zira, David, plus the newer natural voices in Edge), on macOS and iOS the Apple voices (Samantha, Daniel, and the enhanced ones you can download in Settings → Accessibility), on Android the Google voices, and Chrome adds its own set of Google voices that need a network connection — marked online here. Quality varies enormously: install the enhanced or natural voices for your language in the OS settings and this page uses them automatically.
Why the text is read in pieces
Chrome silently stops speaking after roughly fifteen seconds of a single utterance — a long-standing bug — so a whole article handed over in one go just dies mid-paragraph. This reader splits the text at sentence boundaries and queues the pieces, which sidesteps the bug and is what makes the read-along highlighting possible. Very long sentences are split at commas so nothing exceeds the limit.
Uses
- Proof-listening. Ears catch what eyes skim: doubled words, missing words, sentences that never end. Read a draft at 1.2× before sending it.
- Reading fatigue and accessibility. Long documents are easier to absorb by ear, and for dyslexic readers hearing and seeing together helps.
- Language practice. Pick a native voice for the language, slow it to 0.8×, and shadow the pronunciation.
- Multitasking. Listen to an article while doing something else — but keep the tab open; closing it stops the speech.
About saving audio
The browser API plays speech but does not hand back audio data, so a download button is not possible here without sending your text to a server-side voice service — which is exactly what this site avoids. To keep a recording, use the operating system's screen or audio recorder while the page plays. To check how a text reads before it is heard, see the readability checker.
Frequently asked questions
Can I download the speech as an MP3?
No — the browser's speech API plays audio but does not provide the audio data, so there is nothing to save. Recording it requires either a server-side voice service (which would mean uploading your text) or your system's audio recorder while the page plays.
Why are the voices different on my phone?
The voices belong to the operating system, not this page. Windows, macOS, iOS, Android and Chrome each ship their own set. Better voices can usually be added in the system's accessibility or language settings, and they appear here automatically.
The voice stops after a few sentences.
Chrome ends long utterances after about fifteen seconds; this page splits text into sentences to avoid it. If speech still stops, the selected voice may be an online one that lost its connection — pick a voice not marked "online", or try another browser.
Which languages are supported?
Whatever voices your device has — typically 20–60 languages on a desktop. The list is grouped by language code. Choose a voice matching the text's language; an English voice reading French produces gibberish.
Is the text sent anywhere?
Not by this page. Local voices process everything on your device. Voices marked "online" (Chrome's Google voices) send the text to the vendor's servers to synthesise — choose a local voice if that matters.