Don't Scan Blind.
Scan Smart.

QRGuard protects you from quishing attacks phishing links hidden inside QR codes. Before you scan any QR code, let our AI-powered system check it first.

You have 3 free scan(s) left today. Create a free account for unlimited scans.

One Wrong Scan.
Everything Lost.

QR codes look identical whether they lead to a bank website or a phishing trap. Attackers exploit this blind spot embedding malicious URLs in emails, posters, and menus to steal your credentials silently.

99.64% Accurate.
Zero Compromises.

QRGuard's Random Forest model analyzes 24 URL-based features across 100 decision trees delivering a precise phishing verdict in milliseconds, without fetching the page or exposing your device.

Your Shield Against
Digital Deception.

Whether it's a restaurant menu, a payment link, or a promotional poster any QR code can be weaponized. Make QRGuard your first line of defense before clicking anything.

Why QRGuard

Built to Keep You Safe

Everything you need to stay safe from QR-based phishing attacks in one fast, private tool.

ML-Powered
Random Forest model trained on 235,795 URLs from the PhiUSIIL dataset — 99.64% accuracy.
Instant Results
Verdict in milliseconds. No external API calls, no waiting, no tracking.
QR & URL Scan
Upload a QR code image or paste any suspicious URL directly into the scanner.
100% Private
QR codes decoded in your browser. Your data never leaves your device.
Multi-Layer Defense
ML model + GhostBuster + WHOIS intel. Three layers of analysis per scan.
Scan History
Sign up to get unlimited scans and full history. Join free.
Behind the Scenes

How QRGuard Works

Every scan goes through 4 layers of intelligent analysis in under a second.

01
Decode
QR code is decoded in your browser using jsQR — no upload to server.
02
Extract
24 URL features extracted: TLD, length, entropy, keywords, subdomains.
03
Analyze
Random Forest model with 100 trees votes on phishing probability.
04
Verdict
GhostBuster layer checks SSL, domain age, redirects for final score.
Powered by Machine Learning

Trained on 235,795 URLs from the PhiUSIIL dataset

Our Random Forest model uses 100 decision trees and 24 URL-based features to deliver a phishing verdict with 99.64% accuracy in milliseconds, without ever fetching the page.

99.64%
Accuracy
100
Trees
24
Features
0.99
AUC-ROC
Ready to Get Started?

Unlimited Scans. Full History. Free Forever.

Create a free QRGuard account in under a minute. Get unlimited scans, full history, downloadable reports, and direct admin support.

Create Free Account Already a member? Log In