Biometric face recognition compares a live facial image with a trusted reference, such as the portrait on an identity document. Liveness detection performs a separate check to determine whether a real person is present during the verification session.

No. Biometric face recognition checks whether the user’s live facial image matches the portrait on the submitted identity document. Document verification separately evaluates the document, its authenticity signals, and its extracted information.

Face matching should not be used alone to detect deepfakes or spoofing attempts. Active or passive liveness detection adds a separate presence check, while other fraud and verification signals support the final decision.

Create a verification ID securely on your server and open the hosted verification journey through a modal, iframe, browser redirect, or mobile flow. Completed results can then be retrieved through the API or delivered through a configured webhook.

Yes. Users can complete the verification with a desktop camera, move to a mobile device through a QR code journey, or use a mobile-only flow configured for the API key.

okID returns a manual_review status when a verification requires human assessment. Your team can route this outcome into its existing review process before making the final decision.

The final verification status can be verified, rejected, or manual_review. Depending on the configured workflow, the response may also include the verification ID, module results, exported data, custom metadata, and return URL.

The retention period can be configured for each API key. The selected period should reflect the purpose of the verification, applicable legal requirements, and the organization’s data-minimization policy.