For developers · rolling out
Text to Speech API
One endpoint, several engines, credits that never expire. API access is rolling out — in the meantime, every engine it will expose is already testable right here.
Speech studio
Natural voices, ready in seconds
5,000 free characters/day · No sign-up required · Commercial use included
Status: engines live, API rolling out
Being straight with you: the speech engines are in production and serving the web app today, but programmatic API access is not open yet. This page exists so you can evaluate the voices now and get notified the moment keys are available.
Nothing on this page describes an endpoint you cannot call yet — when the API ships, the documentation will replace this section.
What it will look like
The design goal is boring in the best way: a single synthesis endpoint that takes text, a voice, and a format, and returns audio. Engine selection happens through the voice you choose, so switching from a cheap open-weight model to a premium one is a one-field change rather than a rewrite.
- Credit-metered, same balance as the web app — no separate billing to reconcile
- Free-tier engines stay free through the API as well, within daily limits
- MP3 and WAV output
- Keys managed from your account settings, revocable at any time
Evaluate the engines first
The most useful thing you can do before an API launches is confirm the output is actually good enough for your product. Use the generator above with your real copy — not a marketing sample — and compare the free tier against the premium engines. If the free engine already clears your bar, that is the cheapest integration you will find; if it does not, you will know exactly which tier you need.