Detect PII
Reports whether the supplied text contains personally identifiable information, optionally with a per-type entity breakdown.
Authorizations
Bearer authentication header of the form Bearer <token>, where <token> is your auth token.
Body
The workspace is resolved from the API key, never sent in the body.
Text to analyse.
BCP-47 language code. Unset means auto-detect.
Global minimum recognizer score (0.0-1.0). Unset uses the provider default.
When true, the response includes a per-type entity breakdown.
Region codes selecting whole regions of coverage, e.g. "nl", "gb". Every
entity type those regions cover is included, alongside the base catalog.
["all"] is exclusive. Leaving this and entities both empty also runs every
region, so selecting nothing is the widest request rather than the narrowest.
Empty alongside a non-empty entities stays strict. See the capabilities
endpoint for the region list.
Combines with entities: the detector unions the two selections, so a
request carrying both covers the regions plus the named types. Note that
entities only narrows coverage when it travels alone.
The entity types to cover. A named type fires even when it belongs to a
region, so a region's types can be selected individually without naming
the region. Alone this is a strict allowlist; alongside regions it adds
to the region coverage instead of narrowing it.
Per-entity confidence cutoff overrides. An override REPLACES the global
threshold for that type and may sit above or below it: the gateway
lowers the first-pass floor it sends the detector to the smallest value
in the map, and every selected type carries its own cutoff, so a
sub-global override detects MORE of that type without widening the rest.
This only tunes confidence; it never changes which types are covered.
Coverage is decided by entities alone, so every key here must also
appear in entities — a key that does not is rejected, as is any key at
all when entities is empty.
Declarative [0,1] bound mirrors existing constraint on threshold.