BrowserTools

Text to Speech

Speak any text with system voices

By Guga — BrowserTools maintainer, Tbilisi, Georgia Updated 2026-08-18
Free Private No uploads

This page issues no network request of its own — speechSynthesis passes your text to whatever voices the operating system and browser expose. Whether those voices are synthesised locally or by a platform service varies by vendor and is outside this page's control.

What is Text to Speech?

It uses the speech engine your operating system already ships, which means the voice list is whatever is installed locally — a dozen or two on a bare machine, several dozen on a configured desktop, and a different set entirely on a phone. Pick a voice, set rate and pitch between 0.5x and 2x, and press Speak. There is no neural voice here and no download of a model: this is the same synthesis layer that reads a notification aloud. Two limits worth knowing before you paste an article. The page speaks at most the first 5,000 characters of what you typed, and it gives you no way to save the audio — it plays, and that is all. Windows and macOS let you install additional voices in system settings if you want more choice.

Turn written text into natural speech using the voices built into your operating system and browser. Choose voice, speed and pitch — works offline once your voices are installed.

Private Text to Speech In Your Browser

Text to Speech is processed locally with client-side JavaScript. Your files never touch a server, and there is no queue to wait in.

Step-by-Step

How to use Text to Speech

1

Type or paste the passage into the text area; the counter shows how much of the 5,000-character limit you have used.

2

Choose a voice from the picker, which lists the voices installed on this machine grouped by language.

3

Adjust the Speed and Pitch sliders between 0.5 and 2 if the default delivery does not suit the material.

4

Press Speak, and use Pause, Resume or Stop while it is running; there is no download step because no file is produced.

Key Capabilities
  • Lists the voices actually installed on the current device, grouped by language with English first.
  • Offers independent rate and pitch controls from 0.5x to 2x in 0.05 steps.
  • Supports pause, resume and stop mid-utterance without restarting from the beginning.
  • Runs on the browser's built-in synthesis with no model download and no per-use cost.
Common Use Cases
  • A developer proofreading an error message or release note by hearing it read at roughly speaking pace.
  • A language learner checking pronunciation of a passage against a native voice already present on their machine.
  • An accessibility reviewer hearing how an on-screen instruction will sound to a screen reader user.
Clear Answers

Frequently Asked Questions

I typed more than 5,000 characters — will it all be read?
No. Only the first 5,000 characters are handed to the speech engine. The counter will show a number above 5,000 if you exceed it, but the excess is silently dropped, so split a long document into sections and play them one at a time.
Can I download the spoken audio?
No. The speech is a playback event through the browser's synthesis layer, and this page has no access to the audio stream to record. If you need an audio file, use a text-to-speech tool that renders to a file — this one is for hearing text, not for producing a deliverable.
Why are there so few voices, or none I recognise?
The list is exactly what your operating system and browser expose, and that varies widely. Windows and macOS both ship additional voices you can enable in their system settings, and some macOS and iOS voices are downloaded on first use. A bare Linux install may genuinely have only one.
Does my text get sent anywhere?
This page makes no network request of its own — it hands the string to the Web Speech synthesis interface and plays whatever comes back. Whether a given platform's voice is computed locally or by that vendor's service is the operating system's business rather than something this page controls, so treat it the way you would treat your system reader.