A studio for the speech synthesis that is already inside your browser. Pick a voice, shape rate, pitch and volume, watch the spoken words highlight as they play, then copy the exact recipe to reproduce it anywhere.
Text
0 chars0 words— est
Voice
—
1.00×
1.00
1.00
Playback
Voices load asynchronously in most browsers — if the list is empty, wait a second or reload.
The text will highlight here while it is spoken.
Recipe
—
Honest limits: this is the browser's built-in engine (Web Speech API), not a neural voice model — quality depends entirely on the voices your OS/browser provides, and headless or minimal Linux installs may expose none. There is no way to export an audio file from this API; if you need a recording, capture it with your OS while it plays.
Why this exists: hosted text-to-speech is metered per character, and the interesting new open models (Ming omni TTS, Gemini-style voice APIs) all need infrastructure. The boring browser API none of them replaces is still free, offline, and private. Treat this page as the tuning bench for it: find the exact rate/pitch/voice combination, then reproduce it in two lines of code anywhere the Web Speech API exists.