Skip to main content
POST
Remove personal data from a text or a conversation

Authorizations

Authorization
string
header
required

API key as Bearer token

Body

application/json
text
string

One text to redact. Send this or messages.

Maximum string length: 10000
messages
string[]

The messages of one conversation, in order; the same value gets the same placeholder in every message.

Required array length: 1 - 100 elements
Maximum string length: 10000
language
string

Two-letter language code, when known.

mode
enum<string>

protective (default) also removes two kinds of near miss, usually a mistyped real number: a number shaped like a card, IBAN or ID number whose check digit fails, and a phone number one digit longer or shorter than its format allows. precise leaves them. Every other type is removed the same way in both modes.

Available options:
protective,
precise
strict_anonymisation
boolean

Off unless true. Also removes age, school year, supervision, gender, nationality or ethnicity, everyday places and health (English patterns). Safety review reads these, so use it only for text shared after analysis.

Response

200 - application/json

Default Response

text
string
messages
string[]
counts
object

Replaced occurrences per type.

entities_total
number
mode
enum<string>
Available options:
protective,
precise
strict_anonymisation
boolean
model_applied
boolean

Adjustment the API applied to this result, reported for transparency. Descriptive only: route on recommended_action. Absent when it did not apply.

credits_used
number