A real request and the response it produced. Edit the input and run it again — no account needed.
Input
Output
1447ms
{
"classification": "spam",
"reasoning": {
"expectation": "Low",
"value": "Irrelevant",
"red flags": [
"Use of 'URGENT' and 'suspension' creates false urgency",
"Generic 'verify identity' demand lacks context",
"Limited-time offer implies pressure tactic",
"No personalization; generic 'click here' link"
],
"example": "Fake Amazon Prime alerts often exploit fear of suspension to drive clicks, despite no legitimate verification process requiring immediate action."
}
}Call it from your app
curl -X POST https://api.zywrap.com/v1/proxy \
-H "Authorization: Bearer $ZYWRAP_KEY" \
-H "Content-Type: application/json" \
-d '{
"wrapperCodes": ["tc_spam_vs_ham_base"],
"variables": {
"threshold": 0.92,
"sourceText": "URGENT: Your Amazon Prime account is about to be suspended! Click [HERE] to verify your identity immediately. Limited time offer!",
"languageHint": "en-US",
"returnEvidenceSpans": true
}
}'Run this from your own app with 10,000 free credits.
Get a keyEvery one of these has a real example you can run.
Sentiment Classification (3-class)
Classifies text sentiment as positive, neutral, or negative and provides a label plus confidence. Useful for reviews, feedback streams, and social listening.
Academic Integrity Request Detection
Classify whether a student request likely seeks help that would violate academic integrity policies, such as asking for full solutions or ghostwriting. Helps education platforms and institutions flag risky requests.
Support Entitlement Scope Classification
Classify support tickets as in-scope or out-of-scope for a customer support entitlement and, if out-of-scope, into reason categories. Helps support teams apply entitlement boundaries consistently.
Incident Report Business Impact Tier
Classify incident reports into business impact tiers such as minor, moderate, major, or critical based on described effects. Supports consistent severity assignment for incidents and outages.
Sales Lead Fit From Inbound Message
Score inbound sales messages into fit levels such as high, medium, or low based on need, budget, and alignment to the ideal customer profile. Helps sales teams prioritize follow-up and focus on the most promising opportunities.
Upsell and Cross-sell Opportunity Signal
Detect when customer messages indicate potential upsell or cross-sell opportunities and categorize the opportunity type. Helps revenue teams identify expansion signals hidden in everyday conversations.