Free PII redaction tools and references

Everything here runs on the same detector as the API. The playground needs no account; the guides explain what is being detected and where to put the call.

Entity types

Forty types across twelve groups, each with the detection rules and a live demo.

• Email addressesRedact email addresses from text, JSON, and CSV with one HTTP call. Returns exact character offsets and keeps your structure intact.
• Phone numbersRedact phone numbers in national and international formats from text, JSON, and CSV. Model-gated, so dates and ID numbers are not mistaken for numbers.
• Street addressesRedact postal addresses from free text and structured payloads, including partial street names the model only half-captures.
• Postal codesRedact ZIP and postcode values in US 5-digit, ZIP+4, and UK formats from text and tabular data.
• IP addressesRedact IPv4 addresses from access logs, error traces, and analytics exports with a format check per octet.
• Credit card numbersRedact card numbers from text and JSON with Luhn validation, so only real card numbers are flagged and order IDs are left alone.
• IBANsRedact International Bank Account Numbers with mod-97 validation across all SEPA countries.
• SWIFT/BIC codesRedact SWIFT and BIC codes from payment instructions without flagging every all-caps word.

All 40 entity types →

Use cases

• Redact PII before sending text to an LLMStrip names, emails, card numbers, and 37 other entity types from prompts before they reach a third-party model. One HTTP call, no retention.
• Redact PII from CSV exportsSend a CSV and get a redacted CSV back. Per-cell redaction across every column, with headers and structure preserved.
• Redact PII from support ticketsClean customer conversations before they reach analytics, QA tooling, or a third-party summariser.
• Mask PII in application logsRedact IPs, emails, tokens, and device identifiers from log lines before they are shipped to a log aggregator or pasted into a ticket.
• Sanitise production data for stagingTurn a production dump into something safe to load into a non-production environment, without losing the shape of the data.
• Redact data before sharing with vendorsShare a dataset with a contractor or vendor with the personal data removed, and keep an auditable record of what was stripped.
• HIPAA de-identification workflowsRemove names, MRNs, health plan IDs, dates of birth, addresses, and the other HIPAA identifier categories from clinical text.
• GDPR subject access requestsWhen a subject access request covers a shared record, redact the other people in it before you send the response.

Integrations

• PythonCall the Redactly redaction API from Python with requests. One function, exact character offsets, no SDK install required.
• Node.jsCall the Redactly redaction API from Node.js with the built-in fetch, no dependency required.
• curlThe complete curl reference for the Redactly redaction API: text, files, styles, and entity filters.
• n8nAdd a redaction step to any n8n workflow with the HTTP Request node, before a node sends data to an AI model.
• PostmanImport the Redactly API into Postman from its OpenAPI schema and start calling it with a collection variable for your key.
• LangChainWrap a LangChain prompt so personal data is removed before the request reaches the model provider.

Reference

• API referenceEndpoints, schemas, redaction styles, error codes.
• Security and data handlingIn-memory processing, hashed keys, compliance mapping.