Privacy-first, on-device by default
Gemini AI
DSDeepSeek-R1

Stop fake UPI SMS and suspicious links before they trick you

Detect scam messages in real time using DLT sender checks, domain intelligence, and a lightweight rule + ML scoring engine. Works on Android and iOS.

No READ_SMS
DLT sender checks
Link scanner
Hybrid scoring

Why it works

Designed to catch real-world scams without sacrificing privacy or performance.

Real-time SMS analysis

Share to app or use notification listener on Android to flag risky UPI messages instantly.

DLT sender ID logic

Validate registered headers and patterns to spot spoofed or unregistered senders.

Suspicious link detector

Catches lookalike domains, punycode tricks, and shorteners commonly abused in scams.

Privacy-first by default

All checks are offline first. Optional online reputation checks live behind a flag.

Clean REST API

FastAPI backend with sensible rates. Try /v1/analyze with simple JSON payloads.

Hybrid scoring engine

Rule + ML thresholds: Safe <40, Suspicious 40–69, Scam ≥70 for clear decisions.

How it works

Three simple steps to stay safe.

  1. Step 1

    Share or paste the SMS

    On Android, enable notification listener or share from Messages. On iOS, use the Share Sheet or paste.

  2. Step 2

    We analyze safely

    On-device rules check DLT sender patterns and links. Optional online reputation checks can be enabled.

  3. Step 3

    Get a clear verdict

    Scores map to Safe, Suspicious, or Scam with guidance to act. No data selling. No tracking.

Simple REST API

Rate limits: 10 req/min per IP for guests, 60 req/min when authenticated.

Example request

POST /v1/analyze
Content-Type: application/json

{
  "text": "UPI: Your account will be blocked. Pay at http://pay-secure.example.co"
}
Buy me a coffee