Text to Speech

How do I make my browser read text aloud for free?

Paste or type text into AnyTool Text to Speech, choose any voice installed on your device, set the speed, pitch and volume, and press Play. The word being spoken is highlighted as it reads, and long documents are split into chunks so they read to the end. It uses your browser’s built-in Web Speech API, so nothing is uploaded.

Pick from every system voice, grouped by language
Adjustable speed (0.5–2×), pitch and volume — remembered for next time
Play, Pause, Resume and Stop with live word highlighting
Reads long text fully by queuing sentence-aligned chunks
Upload a .txt file or load a sample — 100% client-side, no upload

What is

Text to speech (TTS) is the conversion of written text into spoken audio. In the browser it is provided by the Web Speech API, which synthesises speech from voices installed on the device, with adjustable rate, pitch and volume and word-level boundary events for read-along highlighting.

Text Tools

Related terms

Web Speech APISpeech SynthesisRead AloudScreen ReaderVoice Selection

Frequently Asked Questions

The Web Speech API can cut off or stall on long utterances — Chrome historically stops after a few hundred characters. AnyTool Text to Speech avoids this by splitting your text into sentence-aligned chunks under a safe size limit and speaking them in a queue, one after another. Pause, Resume and Stop continue to work across the whole queue, so articles, emails and book chapters read all the way through.

Some engines, notably Chrome, truncate very long single utterances. This reader splits text into sentence-sized chunks and queues them, so the whole document reads to the end.

Browser speech-synthesis voices are designed to play in real time and the Web Speech API does not expose the audio for saving, so a clean MP3 cannot be produced here. This tool focuses on instant, free, private reading aloud. If you need a downloadable file or higher-quality neural voices, the separate Text to Speech Pro tool is the right place to look.

Not from this tool. Web Speech API voices play live but cannot be exported to an audio file. For downloadable or neural audio, use Text to Speech Pro.

The tool lists all voices reported by your browser, organised into language groups such as English (United States) or French (France). On-device voices work offline; some platforms also offer online voices that need a network connection. A language filter narrows the picker, and your chosen voice, speed, pitch and volume are saved in local storage for next time.

Every voice installed on your operating system and browser, grouped by language, including on-device and online voices. You can filter the list by language.

All synthesis happens locally through the Web Speech API built into your browser. No text is sent to a server, there are no accounts and no tracking, and once the page has loaded it works offline with the voices installed on your device — so you can safely read confidential notes, emails or drafts aloud.

Yes. Speech is generated by your browser, so the text is never uploaded, there is no tracking, and on-device voices keep working offline.

Detailed Explanation

Methodology

How the Text to Speech Reader Works

AnyTool Text to Speech is built on the browser Web Speech API (window.speechSynthesis and SpeechSynthesisUtterance). It enumerates every voice the operating system and browser expose, groups them by language tag, and applies the chosen rate, pitch and volume to each utterance. The current word is highlighted by listening to the utterance boundary event and mapping its character offset back to a precomputed word span in the source text. No audio is generated or processed on a server.

Uses window.speechSynthesis and SpeechSynthesisUtterance — no backend
Lists all system voices, grouped by language, on-device and online
Rate 0.5–2×, pitch 0–2 and volume 0–100% are configurable
Live word highlighting driven by the utterance boundary event
Voice, speed, pitch and volume persist in local storage

How It Works

Reading Long Documents Reliably

Several speech engines, most notably Chrome, truncate or stall on long single utterances. To read full documents the tool splits text into sentence-aligned chunks below a safe character limit, sub-splitting any over-long sentence on clause and word boundaries, and speaks the chunks in a queue. Each chunk records the absolute offset of its first character so boundary events remain correctly mapped for highlighting across the whole document. Play, Pause, Resume and Stop operate over the entire queue.

Sentence-aligned chunking keeps prosody natural at punctuation
A safe per-chunk character cap avoids Chrome long-utterance truncation
Chunk offsets keep word highlighting accurate document-wide
Queue-based playback reads articles, emails and chapters in full

Privacy & Security

Scope, Honesty and Privacy

Web Speech API voices play in real time and the API does not expose the audio stream, so this tool cannot export a downloadable MP3; that requires a separate neural engine, and the interface points users to a dedicated Pro tool for that. Everything else runs locally: text is never uploaded, there is no tracking, and on-device voices keep working offline after first load.

Reading text aloud: in-browser (AnyTool) vs typical online TTS sites
Capability	AnyTool	Typical online TTS
Where it runs	In your browser	On a server
Privacy	Text never leaves device	Text sent to a server
Voices	All system voices, grouped by language	Vendor voice set
Word highlighting	Live, synced to speech	Sometimes
Long documents	Chunked queue reads to the end	Often capped or cut off
Daily character limit	None	Often limited on free tier
MP3 download	No (live playback only)	Sometimes (paid)
Works offline	Yes, with on-device voices	No
Cost / signup	Free, no signup	Often gated or ad-heavy

Server-based TTS can offer neural voices and MP3 export but must receive your text; AnyTool keeps text on-device and reads it aloud instantly using the browser’s own voices.