How do I make a QR code for a bank transfer?

How it works

A QR code stores text in a grid of black and white squares. Phone cameras read the text and act on its form: a web address opens in the browser, a WIFI: line joins a network, a vCard becomes a contact. The codes made here are static: everything is in the squares themselves, so they work without this site, never expire and nobody can count who scans them. So-called dynamic codes hold only a short link to someone's server, which can stop working or record every scan.

The code follows the QR standard ISO/IEC 18004. The text is split into 8-bit codewords, and Reed–Solomon error correction codewords are added so a scratched or partly covered code still reads: about 7 % of it can be damaged at level L, 15 % at M, 25 % at Q and 30 % at H. Larger content or a higher level needs a larger version, from 21 × 21 squares (version 1) up to 177 × 177 (version 40), which holds at most 2,953 bytes at level L. A link such as https://how.box/ is 16 bytes and fits version 1.

The contents follow the formats phones read: the ZXing WIFI: format for networks, vCard 3.0 (RFC 2426) for contacts, mailto: (RFC 6068), SMSTO: for text messages, iCalendar events (RFC 5545), geo: URIs (RFC 5870) and the European Payments Council's EPC QR (EPC069-12) for SEPA credit transfers in euros.

Keep codes short and print them at least 2 cm wide, dark on light. Wi-Fi and event codes depend on the phone's camera app; most current Android phones and iPhones read them.