Text to Speech
Turn text into speech using the voices installed on your device — choose a voice, speed and pitch. Optional online voices are labeled and off by default.
🔒 Only voices installed on this device are listed, so the text stays on your device.
Read any text aloud, right in your browser
Paste or type text, choose one of your device’s installed voices, adjust the speaking rate and pitch, and press play. It uses the browser’s built-in Web Speech API and lists only the voices installed on your device, so the audio is generated there — the text is not sent to a server, and speaking works offline once the page has loaded. Some browsers also offer online voices (Chrome’s Google voices, for one); those appear only if you tick “Also show online voices”, are marked online, and send the text to their provider.
Long passages are automatically split into sentence-sized pieces and spoken back to back, so you can listen to a whole article without it cutting off. While it reads, the sentence being spoken is marked in a read-along box (the word too, where the voice reports its place), ⏮ Previous sentence and Next sentence ⏭ skip around, and a line under the sliders gives the word count and roughly how long it will take at the chosen speed; Volume sits beside speed and pitch. Open file… takes text files and also PDF, EPUB, DOCX and ODT documents, whose text is unpacked for reading. The first time you open a document of one of those kinds, the reader for it is downloaded, so that needs the connection once; speaking the text afterwards does not. Available voices depend on your operating system and browser; installing extra language packs adds more voices, including many languages beyond English and Korean. When the voices cover several languages, a Language menu narrows the voice list to one of them, each named in the page’s language with its count of voices. Tick “Read a script in several voices” and lines that start with a name and a colon — “Ann: Hello” — are read by their speakers, each with a voice of its own, and the names are not spoken. Go to sentence picks any sentence to start from, or jumps there while it reads. Tick “Keep this text and where the reading stopped” and the text and the sentence reached are saved in this browser: they come back on your next visit without playing, “⏯ Continue from sentence …” picks up there, and unticking deletes the saved copy.
Frequently asked questions
Is the text uploaded anywhere?
No. Only voices installed on your device are listed, and speech is synthesized there — the text does not leave your computer. If you tick “Also show online voices” and pick one marked online, the text is sent to that voice’s provider, and the page says so before you press Speak.
Why are only some voices listed?
The available voices come from your operating system and browser. Installing additional language or voice packs in your OS settings adds more options to the list. Voices your browser fetches from the internet are hidden unless you tick “Also show online voices”.
Can I download the audio as an MP3?
Not directly — the Web Speech API plays audio but does not expose a downloadable file. You can use your system’s audio recorder to capture it if you need a file.
Does it work offline?
Yes, once the page has loaded, because the voices listed are the ones installed on your device. No internet connection is needed to speak the text — unless you turn on online voices and choose one. Opening a PDF, EPUB, DOCX or ODT file is the other exception the first time: the reader for that format is downloaded then, which needs the connection once.
Can it follow along as it reads, and read a PDF or an e-book?
Yes. While it speaks, the sentence being read is marked in a read-along box, and the word too where the voice reports its place; “⏮ Previous sentence” and “Next sentence ⏭” skip back and forth, and a line above the read-along box shows which word and sentence it has reached. “Open file…” takes PDF, EPUB, DOCX and ODT documents as well as text files and reads out their text; a scanned PDF with no text layer, or a locked one, is refused.
Can I narrow the voice list to one language, or read a dialogue in different voices?
Yes. When your voices cover several languages, a “Language” menu above the voice list narrows it to one — “All languages” is the default — and picks a voice installed on the device in that language where there is one. Tick “Read a script in several voices” and each line that starts with a name and a colon, like “Ann: Hello”, is read by its speaker: every speaker gets a “Voice for …” menu, starting on the installed voices in the main voice’s language in turn, and the names are not spoken. A time or a web address is not taken for a speaker, speakers past the number the page names are read in the main voice, and speed and pitch stay the same for the whole reading.
Can it remember the text and where I stopped?
Yes, if you ask it to. Tick “Keep this text and where the reading stopped” and the text, with the sentence the reading had reached, is saved in this browser rather than on a server. On your next visit it comes back, and nothing plays until you press a button: “⏯ Continue from sentence …” picks up where you stopped, and Speak starts from the top. A very long text is read but not kept, and the page says so. Untick the box to delete the saved copy. “Go to sentence” picks any sentence to start from, or jumps to it while reading.
Can I use the audio commercially?
This page does not make an audio file: it plays your text through the voices installed on your device (or an online voice, if you choose one). What you may do with a recording of a voice is set by the voice’s maker — Microsoft, Apple, Google or whoever made the voice pack — not by this site, so check their terms before publishing one.