Round QR code generators compared — which actually scan
A real shootout of round QR code generators with print-test data — most produce codes that pass on screen and fail in the printer. Here is the honest list.
Most "best round QR code generator" lists are decoration. The tools they rank produce codes that look great on the preview screen and fail under a printer. The decoder doesn't care that the modules are aesthetically round — it cares whether each module covers enough area, whether the three finder patterns are recoverable, and whether the centre logo left enough error-correction budget to absorb whatever else goes wrong. Most generators get at least one of those wrong.
This post is the shootout. Eight tools tested under the same conditions, the same payload, the same printer, the same scan distance, the same three phones. Pass means a code that scanned reliably from 30cm on every phone. Fail means a code that scanned in the preview and missed at least once on print. The pattern is consistent enough that you can predict it from the generator's UI before testing, once you know what to look for.
A small grid of round QRs done right — what the output should actually look like before we get into which tools fail to produce it. One step further along the "looks great, sometimes scans" spectrum sit the animated QR codes that earn their scan-rate cost only on specific surfaces — same trade-off as round modules, more extreme. Want to try the round-modules call yourself before reading the shootout? The live QR designer renders the same options the tools below claim to ship, with no signup gate in front of the export.
What a round QR code generator has to do
A round QR generator has to solve three problems the square version doesn't.
Module area compensation. A round module covers about 78% of the cell its square counterpart fills (π/4 of the bounding square). Lose a fifth of the dark-vs-light contrast on every dark module and the decoder's sampling step gets noisier. The fix is oversizing each round module to roughly 105% of the cell width, so the circles overlap slightly. Generators that don't oversize produce codes that scan at 5cm and miss at 30cm.
Finder-pattern handling. The three corner patterns are how the decoder locks onto the code. They have to be recognisable as squares (or near-squares with gently rounded outer edges). Generators that round the finders into concentric circles produce codes that look avant-garde and fail decoder recognition entirely. Apple's Camera app is particularly strict about this since iOS 15.
Logo footprint and error correction. Round modules already lose 22% area per module. A centre logo eats further into the error-correction budget. The same logo size that works at level H on a square QR breaks the same code at level Q with round modules. Generators that don't bump error correction automatically when round modules are picked produce codes that scan at level H and quietly fail when the user picks level M. The recalculated ceilings — roughly 14–18% logo coverage at level H, 11–14% at Q, and lower below — sit in the round-QR-with-logo centre size math.
The honest test isn't whether the tool offers a round-modules option. The honest test is whether the tool oversizes the modules, keeps the finders square, and bumps error correction to compensate for the logo. Those three things separate the generators that produce print-ready output from the ones that produce screenshots.
How the shootout was run
Same payload across all tools — a 28-character HTTPS URL with a single path segment. Same physical print size — 30mm × 30mm on satin photo paper at 600 DPI. Same scan distance — 30cm. Three phones — iPhone 14 Pro, Pixel 8, Samsung Galaxy A54. Three lighting conditions — daylight near a window, indoor LED at 500 lux, dim indoor at ~150 lux. Each test ran ten scan attempts; pass means ten of ten on every phone in every condition.
Round modules with a centre logo, error correction set to whatever the tool's default was for round + logo, no other style changes. The aim was to test the tool's defaults the way most users actually use it — pick the round option, drop in a logo, click download. Most users don't crank up error correction manually.
The shootout
Three of the eight produced print-ready output at the default settings. Two were flaky — they passed in daylight and missed at least once in dim indoor. Two failed outright. One passed but only because it didn't offer the centre-logo feature that breaks the others.
The pattern across the failures is consistent. The tool offers a round-modules toggle as a style option without doing the engineering underneath. The output looks correct in the SVG download because SVG handles fractional pixels gracefully. The PNG output from the same tool, rasterised at print resolution, loses the slight overlap between adjacent modules that the SVG had — and the decoder gets a noisier signal from the print.
QR Code Styling — the open-source baseline
QR Code Styling is the JavaScript library most paid generators depend on under the hood. It oversizes round modules by 5% by default, keeps finder patterns square, and exposes the error-correction level for manual control. It produces a working PNG at print resolution with no further tuning.
If you're a developer building your own generator, this is the library to start with. If you're a non-developer using a paid generator, ask the vendor whether they use QR Code Styling or built their own renderer — the answer correlates strongly with whether the output prints cleanly.
Beaconstac — the closest commercial equivalent
Beaconstac (rebranded to Uniqode in 2024) is one of the few paid generators where the round-module output prints cleanly at default settings. It oversizes modules, keeps finders square, and auto-bumps error correction when a centre logo is added. The pricing is per-month subscription with caps on dynamic-QR count, which is a separate question from whether the codes scan — for the question this post answers (do round codes from this tool actually print well?), Beaconstac passes.
The trade-off: dynamic codes live on Uniqode's domain by default, so every QR points at qr.uniqode.com/abc unless you upgrade to the custom-domain tier. That's a domain-trust question, covered in branded short links and the click you lose.
Linked.Codes — the disclosure
I run Linked.Codes. The platform produces round QR codes with 108% module oversizing, square finders, and automatic error-correction bump to Q or H whenever a centre logo is added. Print-test: passes the same shootout. The codes I'm shipping pass the same bar I'm holding everyone else to.
The disclosure isn't false modesty — there are still cases where Linked.Codes isn't the right pick. If you only need one static QR for a one-off event, free vs paid QR code generators covers when free is genuinely the right answer. If you're a developer who wants to build the rendering yourself, QR Code Styling is the right starting point. The Linked.Codes case is the same as the Beaconstac case — you want dynamic round QRs on a domain you control, with analytics, with editable destinations after print, on a single one-time payment instead of a subscription.
If round QRs that print clean on the first try matter for your work, the Linked.Codes lifetime tier ships them by default — module oversizing, square finders, auto error-correction, and a custom domain for the redirect.
See the lifetime tierScore any generator before you commit
The fastest way to evaluate a generator you're considering is to run a small, honest checklist on the round-QR output it produces. The widget below scores any tool against the same five criteria the shootout used. Mark what the tool does, get a verdict.
Round QR generator scoring sheet
Run this on whichever tool you're considering before you commit a print run. A score of 4/5 or 5/5 is the threshold for production use. 2-3 is decorative only. Below that, find a different tool.
What about "decorative-first" generators
A small category of generators sells "artistic QR codes" — codes shaped like animals, woven into illustrations, embedded in landscape photos. Most of them produce output that scans in their own preview and fails on print. The reason is that their decoder happens to be unusually permissive (often a custom one shipped as a JavaScript demo) and real-world scanners are stricter.
If you want a genuinely creative round QR with a recognisable image inside it, the path is QR Code Styling with a custom render hook plus manual error-correction tuning. The output goes through the same print test as a normal round QR. There's no shortcut.
Round QR codes versus square — when round is wrong
This post is about generators, not about whether to pick round at all. The shorter answer to that question: round modules are correct when the brand or the surface calls for them, and wrong when the print quality, the lighting, or the scan distance is uncertain. The full case is in round QR codes — what makes them work. The shape comparison is in custom QR code shapes. The error-correction trade-off is in QR error correction levels.
If you're choosing between square and round purely on aesthetics, square is more forgiving and you'll spend less time on the print-test. Round is worth the extra care when the QR is part of an established visual system that already uses circular elements — a centre-monogram brand, a Spotify-circle aesthetic, a cosmetics-pack with circular bottles. Outside those cases, square is the safer pick.
What to look for if you're picking a tool today
Five questions to ask any round QR generator before you commit.
Does it oversize the round modules? Download a sample at high resolution and zoom in. The dark circles should slightly overlap their neighbours. If they don't touch, the generator didn't compensate, and the codes will print flaky.
Are the finder patterns square? Look at the three corner blocks. Square or near-square with gently rounded outer corners is fine. Concentric circles is broken. This is the fastest visual tell that the generator confused "make it round" with "round everything indiscriminately".
Does error correction auto-bump when you add a logo? Add a centre logo and check what error-correction level the tool uses. If it's still L or M with a logo, the budget is exhausted and the code will be flaky.
Is the output at least 600 DPI when you choose print sizes? SVG sidesteps this question. PNG must be 600 DPI minimum at the print size you'll use.
Does the dynamic version use a domain you control? Not strictly a print question, but a long-term operational one. If the codes redirect through the vendor's domain, you're locked into the vendor's continued operation. If you can point the codes at your own domain, you're not.
A generator that hits all five is rare. A generator that hits at least four is print-ready for most cases. Below four, find another tool.
Related reading
- Round QR codes — what makes them work — the design rules behind why module oversizing matters.
- Custom QR code shapes — what scans and what doesn't — the broader shape decision-frame round fits inside.
- QR error correction levels — when each is right — the L/M/Q/H trade-off this post depends on.
- Free vs paid QR code generators — what actually changes — when free is genuinely the right answer.
- QR codes — platform docs — how the dashboard wires every QR feature, including round.
Why does my round QR scan in the preview but not the print?
Three usual reasons. The modules weren't oversized, so adjacent dark circles don't touch and the decoder loses contrast. The PNG was rasterised below 600 DPI for the print size. Or the centre logo is bigger than the error-correction budget allows. Run the scoring sheet above on the tool — at least one of the three will fail.
Can I just use Adobe Express for round QRs?
For screen-only use, yes. For print, no. Adobe Express does not oversize round modules at default settings, and the print output is flaky in dim indoor lighting. Use QR Code Styling, Beaconstac, or a similar tool that handles the print-engineering layer.
Is there a free generator that gets all five right?
QR Code Styling, the open-source library, hits four of five at default settings. The fifth (custom domain for dynamic redirect) is not in scope for a static-QR library. If you only need static round QRs, the library produces print-ready output for free.
What error correction level should round QRs use?
Q (25%) at minimum, H (30%) if you have a centre logo. Round modules already lose 22% area each, so the error-correction headroom matters more than on square. L and M are too tight for round + logo combinations.
Will my code work on every phone?
Print-ready round QRs scan on every iOS 11+ and Android 9+ camera app, plus every dedicated scanner app worth using. The phones don't differ much; the generators do.
Does the round-modules toggle even matter on small QRs?
It matters more on small QRs, not less. The smaller the print, the smaller each module, and the smaller the contrast budget. A flaky round generator that passes at 50mm side length will fail at 25mm.
Should I pick a tool by features or by output quality?
Output quality. Feature checklists let bad generators look complete. Print-test is the only honest filter. Run the scoring sheet on whatever you're considering before you put the codes on anything you can't reprint.
Sourcesshow citations
- ISO/IEC 18004:2024 QR code bar code symbology specification — https://www.iso.org/standard/83389.html
- QR Code Styling library on GitHub — https://github.com/kozakdenys/qr-code-styling
- Apple iOS Camera scanning docs — https://developer.apple.com/documentation/visionkit/recognizing-tables-within-a-document
- Google ML Kit barcode scanning reference — https://developers.google.com/ml-kit/vision/barcode-scanning
- WHATWG / W3C on URL handling for QR-encoded URLs — https://url.spec.whatwg.org/
- Print resolution guidance for small-format codes — https://www.adobe.com/creativecloud/design/discover/print-resolution.html
Try it on your own domain
Branded short links and dynamic QR codes, on your subdomain or your own domain. One-time purchase, no per-click fees.