Verify your source
Drop in a raw entropy sample. We run the public NIST-style health tests right here in your browser so you can watch the standard clear your source with your own eyes. These are the ordinary tests — the floor every source beats. They are not what we sell. What we sell is finding what passes them.
The public floor — not our deep scanAny binary file works — raw TRNG output, before conditioning. No format restriction, no keys or secrets, just noise.
Nothing is uploaded. Your file is read and analysed entirely on this device. It never reaches our servers.
The standard reads min-entropy from a small slice. At that size the estimator's confidence band is wide — a real deficit and pure noise look the same. More data narrows it until the truth separates from the noise. Closing that gap, over the whole stream and at a controlled false-alarm rate, is what our deep scan is built for.
This is the public NIST-style baseline (RCT, APT, MCV min-entropy), with conservative one-shot cutoffs — the floor, not our deep scan. Your data was analysed locally and never left this device.
What the standard cleared — and we didn't
You just watched the standard clear a source. Here is a real one we measured: 10 units of a Nordic nRF52840 on-chip TRNG, 24 hours, ~13 GB of raw output, 0 integrity errors. Every unit passed the NIST SP 800-90B battery. This is what our deep scan found in them — at the same ~0% false-alarm rate as the tests that cleared them.
Every figure is measured at the same ~0% false-alarm rate as the standard tests, confirmed on a null control, and reproducible from the signed evidence pack. Measured on the FIT IoT-LAB open testbed. We show you the result and the proof — never the detector.
Send us a raw sample and what you need to prove. We scope it under NDA, with encrypted transfer and a defined retention and destruction policy.