Settings
Configure which types of PII are detected and how the app behaves.
Detection Engine
Controls how PII is identified in your text
NER Model (names, orgs, locations)
Uses a local BERT model (~30 MB, cached after first load) to detect unstructured PII like person names, organizations, and locations.
Auto-mask on paste
Automatically run detection when you paste text into the Mask editor, without requiring a click.
PII Categories
Toggle which types of personal data are detected and masked. 13 / 13 active
Identity
<PERSON_1>Full names, first names, last names
<ORG_1>Company, institution, and team names
<LOC_1>Cities, countries, landmarks
<ADDRESS_1>Street addresses with house numbers
Contact
<EMAIL_1>Email addresses in any format
<PHONE_1>Phone numbers, including international formats
Government IDs
<SSN_1>US Social Security Numbers (XXX-XX-XXXX)
<DOB_1>Dates that appear to be dates of birth
Financial
<CC_1>Credit/debit card numbers with Luhn validation
<IBAN_1>International bank account numbers
Technical
<API_KEY_1>API keys, tokens, and secrets (sk-, ghp_, AKIA…)
<CRYPTO_1>Bitcoin, Ethereum, and other wallet addresses
<IP_1>IPv4 and IPv6 addresses
These settings are stored only in your browser's local storage — they contain no PII and are never transmitted anywhere.Detection and masking always happen 100% in your browser.