London, April 10, 2025 – Keyless, the leader in privacy-preserving biometrics, has been ranked 11th out of over 150 vendors in the latest NIST FRVT 1:N evaluation, achieving 99.93% accuracy when identifying faces from a database of 1.6 million high-quality user identities.

The results follow NIST’s Face Recognition Vendor Test 1:N evaluation, a leading global benchmark that tests how accurately algorithms can identify individuals from large datasets.

Unlike the 1:1 test Keyless took part in earlier this year which focused on verifying one-to-one face matches (e.g. login scenarios), the 1:N benchmark simulates real-world identification use cases – such as airport security, border control, and forensic investigations – by testing how accurately a system can pick the correct person out of millions.

Keyless was evaluated across multiple challenging scenarios, including Mugshot-to-Mugshot (high to high quality) and Mugshot-to-Webcam (high to low quality). With high-quality images, Keyless correctly identified the right face 99.93% of the time across a database of 1.6 million – ranking 11th out of more than 150 vendors.

“These results show we can deliver high accuracy even when image quality is poor or the system needs to be strict about false matches,” said Paolo Gasti, Co-Founder and CTO of Keyless. “We’re especially proud of our strong performance in the webcam scenarios, which better reflect the real world.”

Keyless stands out as one of the only vendors in the evaluation to use Zero-Knowledge Biometrics™ (ZKB) – a cryptographic approach to biometric authentication that ensures:

These results reinforce that security and privacy don’t have to be a trade-off, with Keyless’ ZKB technology now trusted by leading banks, fintechs, crypto exchanges, and gaming companies globally.