Account

What new tool do you want?

Generators

QR Code Scanner & Safety Preview

Scan a QR code from an image or your camera — and preview where it leads before you open it, with warnings for suspicious links.

How to use

  1. 1Upload or paste a screenshot of a QR code, or use your camera.
  2. 2Check the decoded content and link warnings before opening anything.

Frequently asked questions

Is it safe to scan unknown QR codes?

QR codes can hide phishing links ('quishing'). Previewing the destination first — as this tool does — lets you spot fake or shortened links.

Is my image uploaded?

No — decoding runs entirely in your browser.

Why won't my code scan?

Try a sharper, well-lit image with the whole code visible and some white margin around it.

Two ways in, one decoder underneath

Whether you upload an image, paste a screenshot (Ctrl/Cmd+V) or scan live with your camera, the same decoding step runs every time: the image is drawn to a canvas, downscaled if it's larger than 1600px on the long edge, and handed to the jsQR library with both normal and inverted-color detection attempted. For the camera path, that same decode loop runs on every animation frame against the live video feed until it finds a code or you stop it — essentially a continuous burst of single-frame scans rather than specialized video tracking.

Why a blurry or angled photo fails when a direct scan wouldn't

jsQR looks for the QR code's three distinctive corner squares (the 'finder patterns') and the timing pattern between them to figure out the code's orientation and grid — motion blur, a steep viewing angle, glare, or a code that's too small relative to the image resolution all degrade those patterns enough that detection fails outright rather than partially succeeding. There's no partial credit: either the geometry can be recovered cleanly enough to read the modules, or the tool reports no code found. A sharp, well-lit, head-on photo with a plain margin around the code is the single biggest factor in a successful scan.

The safety analysis is pattern-matching, not a verdict

After decoding, the tool classifies the content (Wi-Fi, contact card, email, phone/SMS, website or plain text) and, for links specifically, flags several known red flags: unencrypted http://, known URL shorteners that hide the real destination, raw IP addresses instead of domain names, punycode domains that can visually imitate real brands, lookalike spellings (like 'paypa1' or 'g00gle'), an '@' hiding the real host, suspicious TLDs, and login/verify/payment-sounding paths. This is heuristic pattern matching against a fixed set of known tricks — it catches common 'quishing' (QR phishing) setups but is not a security guarantee, and the absence of warnings only means none of the known patterns matched, not that the link is confirmed safe.

Example: previewing before trusting a flyer's QR code

Photographing a QR code stuck on a parking meter and decoding it might reveal `http://paypal-secure-verify.xyz/pay` — the scanner would flag this with three separate warnings: unencrypted http://, a brand-imitating hostname ('paypal' combined with 'verify'), and an unusual .xyz domain ending often used in scams. Seeing all three before tapping 'Open link' is the entire point of previewing: a code stuck over a legitimate parking payment sticker is a documented real-world scam, and this step catches it without having to visit the page first.

Related tools

More: Generators