HomeGuides › PII Redaction API: Strip Emails, Phones, SSNs from Any Text

PII Redaction API: Strip Emails, Phones, SSNs from Any Text

Support tickets, chat transcripts, and LLM prompts routinely contain emails, phone numbers, SSNs, credit cards, and IP addresses.

Redacting before you store or forward text keeps personal data out of logs, analytics, and third-party models.

Redact in one call

curl
curl -X POST "https://pii-detector-redactor.p.rapidapi.com/redact" \
  -H "X-RapidAPI-Key: YOUR_KEY" \
  -H "X-RapidAPI-Host: pii-detector-redactor.p.rapidapi.com" \
  -H "Content-Type: application/json" \
  -d '{"text":"Email [email protected] or call 555-123-4567"}'

Response

counts breaks down what was found by type, so you can alert or block on specific categories.

Response (200)
{
  "redacted": "Email **************** or call ************",
  "counts": { "email": 1, "phone": 1 },
  "total": 2
}

Detect without redacting

POST /detect returns each match with its type and position for custom handling. POST /analyze returns a risk summary for a piece of text.

LLM pre-processing

Run user input through /redact before sending prompts to a third-party model. Personal data never leaves your boundary, and the model still gets the structure it needs.

Run it in production

PII Detector & Redactor has a permanent free tier — 1,000 requests a month, no credit card. Paid plans start at $5/month for 100,000 requests.

Coming soon on RapidAPI ↗

Full endpoint reference →

FAQ

What types of PII are detected?

Emails, phone numbers, SSNs, credit cards, and IP addresses, with pattern-based detection for each category.

Is the submitted text stored?

No. Text is processed in memory and discarded once the response is sent.

Can I customize the masking style?

Yes. The redact endpoint supports mask, label, and remove modes, plus a custom maskChar.