Names nicknames
Heuristicnames_nicknamesBest-effort — fuzzy, approximate, or inferred; check the confidence score.
Bidirectional English name ↔ nickname lookup. Pass 'Robert' to get its nicknames, or 'Bob' to get formal candidates. Ambiguous nicknames ('Rick' → Richard / Eric / Patrick) return a ranked list. Used for record matching when one system stores 'Bob' and another 'Robert'. Processing multiple inputs? Use `names_nicknames_batch` to do up to 1000 in one call.
Parameters
namestringrequiredCall it
curl -X POST https://api.snipget.ai/v1/names/nicknames \
-H "Authorization: Bearer sk_live_..." \
-H "Content-Type: application/json" \
-d '{"name":"Bob"}'The Batch tab sends up to 1000 inputs in one call via /v1/names/nicknames/batch.
Example response
{
"input": "Bob",
"direction": "nickname_to_formal",
"canonical_formal": "Robert",
"candidates": [
"Robert"
],
"is_ambiguous": false,
"is_valid": true
}FAQ
What does this solve?
One system stores 'Bob' and another stores 'Robert', and you need to know they are the same person. Give it either form and it returns the counterpart, drawing on a maintained list of common English name and nickname pairs you should not have to guess at.
How accurate is the answer?
This is best-effort, not a guaranteed fact. A nickname can map to more than one formal name, for example 'Rick' could be Richard, Eric, or Patrick. When that happens is_ambiguous is true, canonical_formal carries the most common pick (Richard for Rick), and candidates lists every possibility alphabetically - use canonical_formal as the default, not candidates[0].
Which way does it work?
Both ways. Send a formal name like 'Robert' to get its nicknames, or send a nickname like 'Bob' to get the formal candidates. The direction field tells you which way it matched.
What does it NOT cover?
It is English-only. Spanish, Arabic, East Asian, and other naming traditions are out of scope. It also only knows well-attested pairs, so unusual or invented nicknames will not be found.
What if the name is not in the list?
You get is_valid false with a short reason and confidence 0, not an error. That means the name simply is not in the catalog, so do not treat a miss as a failure.