DocomintTexto para fala
Converta texto em fala natural online e grátis: ouça uma prévia instantânea com as vozes do seu navegador, ou gere um arquivo WAV real para download com vozes neurais offline.
- Loading voices…
21 words • 108 characters • ~0:08 to speak
What is a Texto para fala?
A free online text reader with two modes, both running entirely in your browser. Instant mode is TTS in the classic sense: press play and your text is read aloud right away using the voices already installed on your device — no download, but no file to keep either. Neural mode is a real offline text-to-speech voice generator (Kokoro, via WebAssembly): after a one-time voice-model download it turns your text into an actual .wav file you can play, keep, and download, instead of a live-only playback. Either way the text never leaves your device to generate the audio — there's no account, no word limit, and no charge for either mode.
How to use the Texto para fala
- Type or paste the text, or press Try an example
- Pick a voice, language, and speed — Instant for a quick listen, Neural for a downloadable file
- Press Preview Audio to listen now, or Generate Audio to create a real audio file to download
Example
Pasting a paragraph, picking the Heart voice, and pressing Generate Audio produces a real downloadable .wav clip of that paragraph read aloud — reusable in a video, podcast, or presentation, unlike a live-only browser preview.
Why use Texto para fala
Real audio download
Neural mode generates an actual .wav file — download it, not just a live-only browser playback.
Named voices with real gender/accent info
Browse offline neural voices by gender and US/UK English accent, with a per-voice Preview button before you commit.
Instant preview, zero download
Prefer to just listen? Instant mode reads text aloud with your device's own voices, with no wait and no file.
Speed, pitch, and volume control
Adjust playback speed on both modes, plus pitch and volume for Instant mode's native voices.
Live word count & duration estimate
See word/character counts and an estimated speaking time update as you type.
Nothing uploaded, either way
Instant mode speaks locally (unless you pick a network voice); Neural mode's model runs and stays entirely on your device.
What can you create?
Video & podcast voiceovers
Generate a real audio file for a script, ad read, or narration track.
Proofreading by ear
Hear your writing read aloud to catch awkward phrasing before you publish.
Language learning
Listen to how a sentence is actually pronounced in a chosen accent.
Accessibility
Have any text read aloud for low-vision users or anyone who prefers listening to reading.
Study & content review
Turn notes or articles into audio you can listen to while doing something else.
Text to Speech vs Speech to Text
| Text to Speech | Speech to Text | |
|---|---|---|
| Direction | Text → audio | Audio → text |
| Typical use | Voiceovers, proofreading, accessibility | Dictation, transcription, note-taking |
| Real-time mode | Instant preview (native voices) | Live microphone dictation |
| Offline neural mode | Generate Audio → downloadable .wav | Upload Audio/Video → editable transcript |
Frequently asked questions
Can I download the speech as an audio file?
Yes, in Neural mode — press Generate Audio and it produces a real .wav file you can play and download, generated on your device by an offline neural voice model. Preview Audio (Instant mode) is for quick, live-only listening using your browser's own voices and doesn't produce a file.
Is my text sent anywhere to generate the speech?
Neural mode's voice model runs entirely on your device — your text never leaves your browser to generate that audio. Instant mode depends on the voice you pick: voices your OS ships locally speak entirely offline, but some browsers also list network voices (Chrome's "Google" voices, for example), and picking one of those does send the text to that provider's server to synthesize the audio, the same as it would in any other app using them.
Why do the available voices look different on my phone versus my laptop?
The voice list comes from your operating system and browser, not from this site — so it varies by device the same way it would in any other app using system text-to-speech, like a screen reader.
How big is the neural voice model, and does it download every time?
It's roughly 90MB, downloaded once the first time you press Generate Audio or preview a neural voice. Your browser caches it after that, so later visits reuse the cached copy instead of downloading it again.
What languages does Neural mode support?
Only English (US and UK) voices, currently — that's the full set the underlying model ships. For other languages, use Instant mode, which can use any voice your operating system already provides.
Can I adjust the pitch and volume, not just speed?
In Instant mode, yes — speed, pitch, and volume all adjust the native voice. Neural mode's voice model only exposes a speed control; it doesn't currently support pitch or volume adjustment.
Is this text to speech tool really free, with no sign-up or word limit?
Yes — both Instant and Neural mode are free, with no account and no hard word limit. Very long text just takes Neural mode longer to render into audio; there's no paywall or usage cap on either mode.
Is this different from my device's built-in screen reader?
Instant mode actually uses the same system voices a screen reader would, just through a page you control directly instead of an accessibility tool running in the background. Neural mode goes further: it's a dedicated offline voice model that produces a real, downloadable audio file rather than only speaking on the spot.