Threat console
Text or message
Investigate in seconds.
Paste any message, DM or job offer and get an AI risk verdict instantly. Detect fraud, phishing and social-engineering language.
What would you like to analyze?
AI-poweredBacked by our threat intelligence datasets for sharper accuracy.
Free-form text from any source, chats, posts, emails or notes.
Step 02 · Results
Awaiting inputAnalysis verdict
Run an analysis to surface the live verdict and evidence here.
Ready when you are.
Drop in a URL, message, or file above to populate your analysis verdict here.
- Risk score with confidence band
- Key signals and evidence summary
- Recommended next steps to stay safe
ThreatSnaps analyses are AI-assisted guidance, not legal or financial advice. Always verify through official channels.
About this checker
Most scams now arrive as plain text - a DM, a job offer, a delivery alert, a romance opener or an investment tip. ThreatSnaps reads the language for known social-engineering patterns, urgency cues, impersonation markers and pay-up-front asks, then returns a calibrated risk verdict you can act on.
Red flags to watch for
- Unsolicited message offering high pay for trivial work, or returns far above the market.
- Pressure to move the conversation off-platform to WhatsApp, Telegram or email immediately.
- Requests for an upfront fee, gift card, crypto deposit or your full ID before any service is delivered.
What to do if you were targeted
- Do not pay, do not share ID documents, and do not click any links in the message.
- Screenshot the conversation and report the sender to the platform.
- Block the contact and warn anyone the sender may have referenced as a mutual connection.
Frequently asked questions
What kinds of messages can I analyze?
Any plain text: SMS, WhatsApp, Telegram, Instagram DM, LinkedIn message, dating-app opener, job offer or recruiter pitch.
Do you need the sender's number?
No. Just paste the message body. Including the sender handle improves accuracy but is not required.
Does the analyzer support languages other than English?
Yes. The underlying model handles dozens of languages, though English, Hindi, Spanish, Portuguese and Arabic are the most accurate.