Contributor program · open to users
Voice cloning and online TTS here are free forever, with no sign-up. It stays that way because the people who use it fill in the part they know best: how a dialect word is really pronounced, what a voice is good for, the step a guide left out. Apply, and a person reads it.
The hard part of a free tool is not building it — it is keeping it good. Models change, voice descriptions go stale, and someone always hears when a pronunciation in their language is off. Only daily users catch that. The longer you use it, the clearer your list of what feels wrong — and that list is exactly what we are short of.
One person speaks one language, yet the site has to work in fifteen. That is not something a single person gets right.
No CV needed. Choose a direction and say what you already do.
Contribute clean audio you have the rights to, or describe what a voice is actually good for: narration or dialogue, short video or audiobook.
Can be small: 30 seconds of quiet-room audio, or one better description.
Check the stiff sentences, terms and punctuation on your language's pages. A native speaker spots what machine translation never will.
Can be small: read one page, fix the three lines that feel wrong.
Write the process you actually ran: what you recorded on, where it broke, how you settled the settings.
Can be small: 600 words on one concrete problem.
File a reproducible bug, suggest an improvement, or walk the main flow on your phone and browser.
Can be small: one reproducible error plus a screenshot.
One minute, four fields. No CV — just say which part you want.
Applications are read in order, not auto-approved. If it fits, we reply; if not, we say why.
A concrete task: proof one page, record a sample, reproduce a bug.
Contributed work is credited next to it. Stay longer and you get a contributor record page.
Not in cash at the moment — better to say that plainly. What we can offer: your name on the part you did, early access to new features, and your reports fixed first.
No minimum. A 30-second recording, one page of proofreading, one reproducible bug — each counts as a contribution.
No. Three of the four directions involve no code. Speaking your own language and caring about the details is enough.
Applications are read by a person, in order, usually within days. If nothing comes back, submit again — no need to be shy.
No. Your email is used only to answer this application: not published, not passed to any third party. How you are credited is your call.
Only your own voice, or one you have explicit permission to use. A quiet room and a phone microphone are enough — 30 seconds to 2 minutes beats expensive gear.
