Strip API keys, secrets, IPs, and PII from any text before it reaches an AI tool. Structure preserved. Nothing sent to any server.
Live tool
| Pseudonym | Original |
|---|
Workflow
Paste a config file, log output, code snippet, or .env file into the input panel.
Sensitive values are replaced with consistent pseudonyms. A mapping is stored locally so everything is reversible.
Paste the clean output into any AI tool. Structure is preserved — the AI can still help you.
Paste the AI response into the Deanonymize tab with your mapping. Real values are restored instantly.
Coverage
API keys (AWS, OpenAI, GitHub, Slack), JWTs, bearer tokens, PEM keys, connection strings, hex secrets, SSH keys
IPv4 addresses with CIDR, domain names, email addresses, MAC addresses, internal URLs with ports
Phone numbers (8 countries), credit cards (Luhn-validated), SSNs, passports, IBANs, names, dates of birth
AWS account IDs, ARNs, S3 buckets, Azure subscription IDs, GCP project IDs, cloud resource identifiers
Detects secrets written in 13 languages including French, Spanish, German, Russian, Chinese, Japanese, and Arabic
NHS numbers, bank sort codes, SWIFT/IBAN, driving licences, tax references (UTR, NINO, EIN), vehicle plates, VINs
Every redaction has a mapping. Paste the AI response back and restore all original values with one click.
No servers, no telemetry, no data collection. Runs entirely in your browser or editor.
Get started
Auto-anonymizes when you paste into AI chat sites. Auto-restores when you copy the response. Works on 30+ AI tools.
Anonymize and deanonymize directly in your editor. Select text, right-click, or use the keyboard shortcut.
Pricing
The core tool is free, forever. Pro and Enterprise plans add pattern packs, custom regex, and compliance support.
Privatiser uses pattern-based detection and may not catch every sensitive value, especially those in unusual formats or custom syntax. Always review the output before sharing. Learn more