ID documents & passports
Passports, national IDs, driver licences and residence permits from around the world, recognised by layout and content.
idMax recognises IDs and passports, MRZ zones, barcodes and credit card numbers, and matches a face against the document. Everything runs locally: no GPU, no cloud round-trip, no personal data in transit.
Runs in the browser via WebAssembly — nothing to install to see it work.
Nearly every document issued in the world — around 3000 types across more than 100 languages — plus the machine-readable data printed alongside them.
Passports, national IDs, driver licences and residence permits from around the world, recognised by layout and content.
ICAO-compliant machine-readable zones parsed and checksum-verified, in every standard line format.
PDF417, QR and 1D symbologies, including the AAMVA payload on North American driver licences.
Card number, expiry date and holder name captured straight from the camera frame.
Compare the portrait on the document with a live selfie to confirm the person presenting it is its holder.
Names, dates, document numbers and addresses returned as typed fields with per-field confidence.
Recognition happens entirely on the device. Document images and extracted data never leave it, so there is no transfer to secure, no third-party processor to audit and no retention policy to negotiate.
The engine runs on ordinary CPUs — office laptops, kiosks, handheld terminals and phones. No accelerator to buy, no driver stack to maintain, no cloud instance to keep warm.
Results arrive in a fraction of a second per frame, with accuracy tuned over years of production use in border control, banking and hospitality.
Native libraries for every major desktop platform, for counter and back-office workstations.
On-device recognition from the camera, offline-capable, for field agents and customer onboarding.
Full recognition inside the browser tab. Nothing to install, and the image still never leaves the machine.
These demos run the WebAssembly build in this tab. Point your camera at a document — the image is processed locally and is never uploaded.
Driver licences and state IDs, including the PDF417 barcode on the back.
Open demoAustralian driver licences and identity cards across the states and territories.
Open demoDemo apps for iOS and Android show the same engine running natively on a device.
One engine, delivered in the shape your project expects. Tell us the target and we send the matching build with a trial licence.
A native library you link into your own application, with idiomatic bindings for each language.
SDKs that read straight from the device camera and keep working with no connection.
The WebAssembly build, loaded by the page itself. Nothing is installed on the visitor's machine.
Whichever build you take, the call is the same three steps: load the engine, hand it an image, read typed fields back. No pipeline to assemble, no model files to manage, no service to keep running.
If on-device processing does not fit your workflow, we also run recognition and document storage on our infrastructure, built to meet current personal data protection standards. On-premise deployment of the same stack is available too.
Discuss your setupTell us which documents you handle and where they are captured — we will point you at the right build and a trial licence.