Developers
Add proof of human to any voice product.
The physics of human speech, and how we measure it.
Quickstart
First mark attested in a few minutes.
API reference
Full endpoint docs.
Verification endpoint
Check a mark programmatically.
SDKs
Drop-in client libraries.
You're building something that listens — a call center, a voice assistant, an agent that takes voice commands. OriginStory gives you a way to know, for any given voice input, whether it came from a live human or not.
It's one API call. Send audio, get back a verified/unverified result, decide what your product does with it. No biometric enrollment flow to build, no new UI for your users, no separate database to maintain. Verification runs on physical signals captured at the point of speech, so the proof exists before the audio ever reaches your servers — you're just asking for the result.