Text to Speech — free TTS in your browser, MP3 download
Two free engines, one tool. Browser TTS uses the Web Speech API built into your browser — instant, offline, every voice your OS knows. High-quality mode renders natural-sounding Amazon Polly voices via the free Puter.js SDK and gives you a downloadable MP3. No login, no quota, no watermark.
High-quality mode renders MP3 audio via Puter.js — a free, no-signup SDK that fronts Amazon Polly's neural voices. Best with prompts under ~3000 characters.
Browser mode plays instantly. High-quality mode renders a downloadable MP3 in a few seconds.
Generating MP3…
Recent MP3 generations
From text to audio in one click
Pick an engine
Browser TTS for instant offline playback with your OS voices. High-quality mode for downloadable MP3 with 29 Amazon Polly voices.
Paste or type your script
The character meter shows the rough audio length at 150 words per minute — handy for narration timing.
Tweak voice, rate, pitch, volume
Browser mode exposes the full SpeechSynthesisUtterance spec. Drag rate up to 2× for skim-listening; drag pitch down for narration.
Hit Speak — or Generate
Browser mode plays immediately with word-level highlighting. High-quality mode calls puter.ai.txt2speech(text, {voice, engine:"neural"}) via the Puter.js SDK (loaded on first use from js.puter.com) and gives you a real MP3 you can download.
What people use TTS for
A free text-to-speech tool isn't just an accessibility helper — it's the cheapest way to prototype voice content before you commit to a paid voice actor.
Podcast intros
Generate a clean MP3 intro / outro you can edit into your show. Royalty-free, no rights issues.
Audiobooks & study notes
Convert chapters or PDFs into audio you can listen to on a commute. Browser mode handles long texts free.
Video voice-overs
Drop the MP3 into your timeline as a placeholder track — or as the final one. Eight distinct voices to choose from.
Accessibility & proofreading
Hearing your draft read aloud catches typos eyes miss. Built right into the browser, works offline.
Frequently asked questions
Yes. Both engines are 100% free with no signup. Browser TTS uses the Web Speech API built into your browser so it's instant and works offline. High-quality MP3 mode uses the free Puter.js SDK, which fronts Amazon Polly's neural TTS voices — no key, no signup, no per-page quota you have to negotiate.
Yes — switch to High-quality MP3 mode and click Generate. You'll get a downloadable MP3 produced by Amazon Polly's neural engine (29 voices across English, Japanese, Korean, Chinese, Spanish, French, German, Italian, Portuguese, Swedish and Russian). Browser TTS mode does not expose the underlying audio stream — that's a Web Speech API limitation, not a HCODX one — so it cannot be recorded or downloaded directly.
Browser mode supports every voice your OS or browser provides — typically 50+ languages including English, Spanish, French, German, Italian, Portuguese, Japanese, Korean, Chinese, Arabic, Hindi and more. The voice dropdown groups them by language. High-quality MP3 mode covers English plus 9 other languages via Amazon Polly voices grouped in the dropdown.
Browser mode never sends text anywhere — synthesis happens entirely on your device. High-quality MP3 mode sends the text to api.puter.com over HTTPS (Puter's SDK fronts Amazon Polly). HCODX itself never sees or stores your text — your browser talks straight to Puter.
Browser mode uses your OS's installed TTS voices, which differ by platform. Windows ships Microsoft voices, macOS/iOS ships Siri voices, Android ships Google voices, and Chrome also bundles some network voices. The High-quality MP3 mode is the same eight voices everywhere because they're rendered server-side.
Browser mode has no real limit in practice — some browsers cap a single utterance around 30,000 characters but you can queue multiple back-to-back. High-quality MP3 mode works best under ~2000 characters per request — for longer scripts, split into chapters and concatenate the MP3s (try our Audio Tools for joining).
About browser-based text-to-speech
The Web Speech API shipped in Chrome in 2013 and is now baked into every modern browser — Chrome, Edge, Safari, Firefox, Opera, Samsung Internet. It exposes the operating system's own text-to-speech engines through a simple JavaScript interface called speechSynthesis, meaning every device you own can talk for free without ever sending text to a server.
Why two engines?
- Browser TTS is instant, private, offline-capable, and has hundreds of voices across 50+ languages — but the audio stream is not exposed by the spec, so it cannot be recorded or downloaded.
- High-quality MP3 mode uses the free Puter.js SDK, which fronts Amazon Polly's neural TTS — 29 natural voices across 10 languages, real MP3 you can ship, free, no key. The trade-off is the text has to leave your browser.
Tips for natural-sounding TTS
- Add punctuation. Commas and periods give the engine pause cues — a comma-less sentence sounds robotic regardless of model.
- Spell out abbreviations if the engine reads them wrong — "Dr." vs "Doctor", "St." vs "Street", "USD 5" vs "5 dollars".
- Drop the rate slightly for narration (0.9–0.95×) — most default voices are tuned for navigation, not storytelling.
- Use the "Surprise me" samples as references — they're short enough to compare voices quickly.
- For Amazon Polly, Brian and Matthew are best for narration, Joanna and Kendra for explainers, Amy and Salli for soft / friendly content.
Browser support
Browser TTS works on all major browsers including mobile Chrome and Safari. iOS Safari requires a user gesture (tap) to start speech — that's a security feature, not a bug. The High-quality MP3 mode works anywhere that can make an HTTPS request.