Text to speech
Type text, hear it spoken in English or Hindi - or generate a voice file to keep. Everything runs on your device.
Listen now uses your device's built-in voice - including Hindi on most phones. Generate audio file runs a neural voice fully on your device and saves a WAV: English voices download about 95 MB the first time (a 63 MB voice model plus the on-device speech runtime) (then work offline until your browser clears site data); the Hindi AI voice is a one-time download of about 160 MB, kept in your browser's cache when it supports caching and has space.
HOW IT WORKS
1
Type your text
English, Hindi, or mixed - a sentence or a full paragraph.
2
Pick a voice
English female or male, or the Hindi AI voice.
3
Listen or download
Hear it right away, or generate a WAV file you can save, share, or convert to MP3.
Questions
What format is the downloaded file?+
WAV - the original, full-quality audio. Need MP3 instead for a smaller file? Convert it with our WAV to MP3 tool.
Why does the first audio file take time?+
The first generation downloads a voice model to your device: about 63 MB per English voice, or about 160 MB for the Hindi AI voice. English models are stored in your browser afterwards, so later uses start right away and even work offline - unless your browser clears site data, in which case they download once more. The Hindi model is kept in your browser's cache when it supports caching and has space, so a later visit may download it again.
Can I download Hindi audio?+
Yes - choose the "Hindi - AI voice" option and tap Generate audio file. This voice runs on the Supertonic 3 model by Supertone under the OpenRAIL-M license; first use downloads about 160 MB, and the use restrictions in our terms apply (you must say the audio is machine-generated when you share it). Hindi listening via Listen now works on most devices without any download. The English voices use public-domain recordings (LJ Speech and LibriVox).
Is my text uploaded anywhere?+
No. Both listening and file generation happen entirely on your device. Your text never leaves your phone or computer.
Do Listen now and the downloaded file sound the same?+
Not exactly. Listen now uses whatever voice your device already has, for speed. Generate audio file uses the neural voice you picked above, which is what ends up in the WAV.