QR Code Module
A QR code module is a single square cell in a QR code’s grid, either dark or light. Each module holds one bit of data or serves a fixed structural role, such as marking a corner. A code contains between 441 modules (Version 1, 21 x 21) and 31,329 modules (Version 40, 177 x 177).
Why QR Code Modules Matter
Modules decide whether a code scans. The scanner reads the grid module by module, so every module must be clearly visible to the camera. Too few pixels per module, a stretched image or a damaged corner all cause failed scans.
The ISO/IEC 18004 standard defines everything else in modules. The quiet zone is 4 modules on each side. Version numbers are module counts. Error correction consumes modules. A QR module is the key to most sizing advice.
QR Code Module vs Pixel
A module is a data cell with no fixed size. A pixel is a screen unit. The two connect through one ratio: pixels per module.
| Term | Meaning | Fixed size? |
|---|---|---|
| Module | One dark or light cell in the QR grid | No |
| Pixel | One dot in a screen or image file | Yes, per display |
| Pixels per module | How many screen pixels draw one module | Set by the designer |
A code drawn at 4 pixels per module renders each cell as a 4 x 4 block of pixels.
How Many Modules Does a QR Code Have?
A QR code has 17 + (4 x version number) modules per side. Version 1 is 21 x 21. Each higher version adds 4 modules per side, up to Version 40 at 177 x 177. The generator picks the version from the data length and the error correction level.
| Version | Modules per side | Total modules |
|---|---|---|
| 1 | 21 x 21 | 441 |
| 3 | 29 x 29 | 841 |
| 10 | 57 x 57 | 3,249 |
| 25 | 117 x 117 | 13,689 |
| 40 | 177 x 177 | 31,329 |
A short link fits Version 2 or 3. A long URL with tracking tags can need Version 8 or higher, which means more and smaller modules at the same printed width.
QR Code Pattern: Types of Modules
A QR code pattern contains five kinds of modules. Four are function patterns that help the scanner. One carries the data.
| Module type | Job | Version 3 count |
|---|---|---|
| Finder patterns | Locate the code and its orientation. Three 7 x 7 squares, each with a 1-module light border | 192 |
| Alignment pattern | Corrects warping. One 5 x 5 square in Version 3 (appears from Version 2) | 25 |
| Timing patterns | Alternating dark and light row and column that set grid spacing | 26 |
| Format information | Stores the error correction level and mask, as 15 bits written twice | 30 |
| Dark module | One fixed dark cell | 1 |
| Data and error correction | The encoded content and the repair data | 567 |
The function patterns use 274 of the 841 modules in Version 3. The data area of 567 modules equals 70 codewords of 8 bits plus 7 remainder bits.
Denso Wave, the company that invented the QR code, built the finder pattern around a 1:1:3:1:1 dark-to-light ratio. That ratio rarely appears by accident in printed material, so scanners can detect it quickly at any angle.
A mask pattern is applied to the data modules. The encoder tries eight masks, scores each for large blocks and false finder look-alikes, and keeps the best one. This is why a QR code looks like random static.
How Many Pixels Per Module Does a QR Code Need?
A QR code needs at least 3 pixels per module in testing, and 4 or more is the safer design target. In a decode test of a Version 3 code (40 trials per setting, simulated camera capture, zxing-cpp 3.1.1 and OpenCV 5.0.0), results were:
| Pixels per module | zxing-cpp | OpenCV |
|---|---|---|
| 1 and 1.5 | 0 of 40 | 0 of 40 |
| 2 | 38 of 40 | 0 of 40 |
| 2.5 | 40 of 40 | 0 of 40 |
| 3 and above | 40 of 40 | 40 of 40 |
The two readers disagree at 2 and 2.5 pixels, so a code that works on one phone may fail on another. Export at whole pixels per module. A 1.5x scale makes some modules 1 pixel wide and others 2, which blurs edges. Phones use their own decoders, so treat these figures as directional.
Which Modules Can a QR Code Lose?
The finder patterns cannot be lost. In the same test, blanking one finder pattern gave 0 of 40 on both readers. Blanking the 25-module alignment pattern still decoded 40 of 40. Blanking the timing patterns or both copies of the format information decoded on zxing-cpp but failed on OpenCV, so a scanner may or may not recover them.
Error correction repairs data modules. It handles one solid patch far better than scattered errors:
- Scattered flips: levels L, M, Q and H held up to about 1%, 2%, 2% and 3% of data modules flipped.
- One solid patch: level H decoded a patch covering 21.3% of the data area. Level L failed at 8.6%.
Blanking a patch sets modules to light, so only the dark ones inside it are actually wrong. Real damage may differ.
Frequently Asked Questions
What is a module in a QR code?
A module is one square cell of the QR code grid, either dark or light. A whole code is a grid of these cells. Modules are a unit of data, not of length, so a module can be drawn with 4 pixels or half a millimetre of ink.
How many modules does a QR code have?
It depends on the version. Modules per side equal 17 plus 4 times the version number. Version 1 has 441 modules in total, Version 3 has 841, and Version 40 has 31,329. Longer data needs a higher version.
Is a QR code module the same as a pixel?
No. A pixel is a screen unit with a fixed size on a given display. A module is a cell of the code with no fixed size. QR code pixels only matter as a ratio: how many pixels draw each module.
What are the three big squares in a QR code called?
They are finder patterns, also called position detection patterns. Each is 7 x 7 modules. They tell a scanner where the code is and how it is rotated. Covering one made the code unreadable on both readers in testing.
Does a bigger QR code have bigger modules?
Only if the version stays the same. Printing the same code larger enlarges every module. A higher version fits more modules into the same width, so each one gets smaller.
To keep the module count low and every module readable, build a short-link QR code with linkutm.