A narrator who never gets tired of take nine.
$0
to re-render the script after an edit
One request returns the whole voiceover track as a finished WAV, read in a voice cloned from ten seconds of you. You rewrite a script more often than you record it, and here you re-render it as often as you want.
Why the meter shows up in the edit, not the invoice
The bill for a ten-minute video is small. The bill for the same video re-rendered eleven times is eleven times that. What the meter buys you is a reason to stop editing.
audio
a scripted essay open, one take, no edit
generated 2026-08-05
English
“Most people get this backwards. They optimise the thing that is easy to measure, and then wonder why the number went up while the product quietly got worse. It happens in every company I have worked in, and it never announces itself, there is no meeting where somebody proposes making the thing worse. There is only a dashboard, a target, and a hundred small decisions that each look reasonable on their own. Let me show you what I mean.”
Leo · 23.4 s · youtube voiceover · scripted essay
What the API does here
- POST /v1/tts/bytes returns one complete WAV, the whole voiceover track in a single request, ready to drop on a timeline.
- The narrator is you. Ten seconds of reference audio, no enrolment step, no per-voice fee.
- Inline controls place the pacing by hand, a break tag where the cut lands, a spell tag for a product code or a URL.
Notes
Can I clone my own voice for narration?
Yes. Ten seconds of your own recorded audio is enough. There is no training job, no approval queue and no per-voice fee.
How long can one script be?
Any length. A single request takes up to 2,000 characters, and long-form work is a loop over those requests, since tokens are unmetered on a stream, a forty-minute script costs the same as a forty-second one.
A key and one stream to build this on, the same production API this page measures.
Get a key for this use case