QR-Code Generator

Configuration options

Two settings decide whether a code scans reliably: how much redundancy it carries, and how large it is printed. ISO/IEC 18004 defines both.

How QR codes work

A QR code stores data in a grid of black and white squares, read in two dimensions rather than one. That is the key difference from a traditional barcode: a supermarket barcode encodes about 20 digits along a single axis, while a QR code uses the whole square and can hold up to roughly 4,300 characters of text or 7,000 digits.

The three large squares in the corners are position markers. They let a scanner find the code and work out its orientation, which is why a QR code reads correctly upside down or at an angle. Smaller alignment patterns keep the grid readable when the surface is curved or the photo is skewed.

Error correction: L, M, Q, and H

QR codes use Reed–Solomon error correction, the same family of mathematics that keeps scratched CDs playable. Redundant data is woven throughout the grid so that a damaged code can still be recovered in full. The four levels trade capacity against resilience:

• L (Low)— recovers from about 7% damage. Most data per pixel; good for clean digital display.
• M (Medium)— about 15%. The usual default and a sensible balance.
• Q (Quartile)— about 25%. Worth choosing for print that will be handled.
• H (High)— about 30%. Necessary if you plan to place a logo over the center, since the logo is simply damage as far as the scanner is concerned.

Higher correction means more modules for the same content, so the code becomes visually denser. If you raise the level, raise the width too or the squares become too small to scan reliably.

Choosing a size

Width here is measured in pixels, from 10 to 512. For screens, anything from 200 pixels upward scans comfortably. For print, the rule of thumb is that the code should be at least one tenth of the scanning distance: a poster read from two meters wants a code around 20 centimeters across.

The margin — the quiet zone — matters more than people expect. Scanners need clear space around the grid to identify where the code ends. The standard calls for four modules' worth of blank margin, and codes that fail to scan when printed have often had that margin cropped away.

Common uses

• Links— the most common case by far. Put a URL on a poster, business card, or packaging without making anyone type it.
• Wi-Fi access— encode network name and password so guests connect without being told the password aloud.
• Contact details— a vCard saved straight to the phone's address book.
• Payments— most mobile payment systems identify a payee by QR code.
• Tickets and boarding passes— a scannable reference that survives a creased printout.
• Inventory and asset tags— link a physical object to its record in a system.
• Restaurant menus— point to a page you can update without reprinting anything.

Making codes that actually scan

• Keep contrast high and polarity correct.Dark modules on a light background. Inverting the colors breaks many scanners.
• Keep the URL short.Less data means fewer, larger modules, which scan more reliably from a distance.
• Preserve the quiet zone.Do not crop to the edge of the grid or let artwork intrude into the margin.
• Avoid curved and reflective surfaces.Glossy stock under bright light can defeat a camera entirely.
• Test before you print.Scan the final artwork with more than one phone, at the distance people will really use.

A note on safety

A QR code is not readable by eye, which makes it a convenient way to disguise a hostile link — a scam sometimes called quishing. Stickers placed over legitimate codes on parking meters and payment terminals are a recurring problem.

Check the URL your phone previews before opening it, be wary of codes in unsolicited post or email, and look closely at whether a code in a public place has been stuck over something else.