FAQ

Onepin is the agent layer for AI voice. Bring a script; we plan it, route each line to the best model, validate every line, fix what fails, and ship publish-ready audio. TTS makes voice. Onepin makes it production-ready.

Every line is scored before it ships for accuracy (WER), naturalness, noise, and pronunciation. Flagged lines are auto-fixed or regenerated. You set the bar. We hold it.

It cleans your script before generation, expanding numbers, dates, abbreviations, and symbols into how they should be spoken (for example, “$20” becomes “twenty dollars”) so the model says the right thing. Nearly half of the TTS errors are fixed by the normalizer.

Onepin already validates every line automatically for accuracy, naturalness, noise, and pronunciation, and fixes what fails before you ship. For a human sign-off on top, human validation is coming soon as a premium add-on.

Yes. Bring your own API key for ElevenLabs and many other providers, run it through Onepin routing and validation, and get noticeably cleaner, more consistent results from the same model.

Use your own TTS provider account inside Onepin. You keep your provider rates and relationship, and we handle routing, validation, and fixing on top. Available on Studio and above.

Not yet, but open-source support is coming soon. Today we route across 30+ commercial models, picking the best one for each line.

Voice cloning is on the way. Stay tuned.

10 languages today, with more coming. Need a specific language or locale? Email support@onepin.ai and we will prioritize it.

Continuously. We benchmark TTS models across languages on a regular cadence, and the results feed straight into auto-routing, so the best model for each line stays current automatically.

We translate multi-lingual scripts through an external translation API, with a premium translation option coming soon.

Seats are free on every plan. Bring your whole team and their agents. You pay for the voice you generate, not headcount.

One currency for the whole pipeline: generation, routing, and validation. 100 credits to the dollar. Each plan lists how many validated minutes that buys.

Included. One credit covers generation and validation together, with no second TTS invoice. We route each line to the best model and pass generation through at cost.

You do not hit a wall. Past your plan, credits meter as pay-as-you-go, or you move up a tier. Free is the only plan without overage.

Not on self-serve plans: each cycle starts fresh. Enterprise agreements are custom, including credit volume and how it is allocated across your workspaces. Talk to us at support@onepin.ai.

Annual billing saves 20%, and you can downgrade to Free or cancel anytime. No lock-in.

Yes, the Free plan lets you try the full validated workflow with 500 one-time credits, no card required. Upgrade to a paid plan whenever you're ready to ship more.

Yes. We assign you every output you generate, and we only route to engines that permit commercial use of what they produce. Enterprise agreements can add a negotiated intellectual-property indemnity on top. Full detail is in our Terms of Service.

Yes. We are SOC 2 Type II compliant and run regular penetration testing.

No. Your data is yours. We train on our own data, never yours.

We are working toward GDPR compliance.

Email support@onepin.ai and a real person will get back to you. Or run your first line free and see the pipeline for yourself.