Text to Speech
Hear your text spoken. Listen instantly using the voices built into your browser and device, in dozens of languages, or generate a natural AI voice as an MP3 file you can download and use in videos, presentations and lessons.
- Private by default
- No sign-up
- Free to use
Uses the voices installed in your browser or device. Voices marked “online” are provided by the browser maker’s servers.
How to use Text to Speech
- Type or paste your text.
- To listen now, choose Listen in the browser, pick a voice, speed and pitch, and select Read aloud.
- To get a file, choose Download as MP3, pick an AI voice and select Create MP3.
- Play the result and download the MP3.
Text to Speech features
Instant listening
Uses your browser’s built-in speech engine with no waiting and no upload for locally installed voices.
Many languages
Every voice installed on your device is listed, with voices for your language first.
Speed and pitch
Slow down for language learning or speed up for proofreading.
Natural AI voices
Nine realistic voices for MP3 output, from calm to expressive.
MP3 download
A standard MP3 file ready for videos, slides and podcasts.
Long texts
Browser reading handles long documents by speaking sentence by sentence; MP3 files cover up to 4,000 characters each.
When to use Text to Speech
- Proofreading: hearing your writing reveals mistakes that eyes skip.
- Creating voice-overs for explainer videos and presentations.
- Listening to articles and study notes while doing something else.
- Supporting reading difficulties, dyslexia and visual impairment.
- Practising pronunciation when learning a language.
Text to Speech FAQ
Is my text sent anywhere?
In browser mode with locally installed voices, speech is generated on your device. Voices marked “online” are produced by your browser maker’s servers. For MP3 files, the text is sent securely to OpenAI’s speech service, and the page tells you so before you create one.
Why are there so few voices in my browser?
Voices come from your operating system and browser. Windows, macOS, Android and iOS let you install more languages in their speech or accessibility settings.
Are the MP3 voices human?
No. They are AI-generated voices. If you publish audio made with them, tell your audience that the voice is AI-generated.
Can I use the MP3 commercially?
Generated audio can generally be used in your own projects, subject to the speech provider’s usage policies. Do not use it to impersonate real people.
How long can the text be?
Browser reading has a generous limit of 20,000 characters. Each MP3 can contain up to 4,000 characters, roughly four to five minutes of speech; split longer texts into parts.
Two ways to turn text into speech
Every modern browser includes the Web Speech API, which lets a web page ask the device to read text aloud. The voices come from the operating system: Windows, macOS, Android and iOS each ship a set of voices and allow more to be installed. Because the speech is produced on the device, it starts instantly and, for local voices, works without sending the text anywhere.
Device voices vary in quality, and browsers do not allow web pages to save their output as a file. For a downloadable recording, this tool offers neural text-to-speech from OpenAI. Neural voices are generated by models trained on recorded speech, so they handle rhythm, emphasis and pauses far more naturally than older synthetic voices.
Listening is one of the most effective proofreading techniques. Reading silently, the brain fills in missing words and smooths over clumsy sentences; hearing the text read literally exposes repeated words, missing articles and sentences that run too long. Slightly increasing the speed helps keep attention while checking long documents.
Text to speech is also an accessibility tool. It supports people with visual impairments, dyslexia or reading fatigue, and helps language learners connect spelling with pronunciation. Choosing a voice that matches the text’s language is important, since a voice reads every text with its own language’s pronunciation rules.