For AI agents: the documentation index is /docs/llms.txt, and this page is available as Markdown.
Search
Turn words into West Kite property IDs: an address, APN, building, owner or company name, with options labelled the way West Kite's search box shows them. Or pass a point and get the parcel under it. Search is free.
GET/v1/search
Parameters
| Name | Type | Description |
|---|---|---|
q |
string | An address, APN, building, owner or company name, such as 550 Madison Ave, New York or Prologis. Up to 200 characters. |
in |
string | With q: a place to prefer, such as San Francisco, CA. |
lat, lng |
number | A point in decimal degrees. Alone, the parcel there; with q, options near it come first. |
limit |
integer | How many results for q. Default 20, up to 50. |
Request
curl --get "https://api.westkite.com/v1/search" \
-H "Authorization: Bearer $WESTKITE_API_KEY" \
--data-urlencode "q=550 Madison Ave, New York"const res = await fetch(
"https://api.westkite.com/v1/search?" + new URLSearchParams({ q: "550 Madison Ave, New York" }),
{ headers: { Authorization: `Bearer ${process.env.WESTKITE_API_KEY}` } },
);
const { data } = await res.json();import os, requests
res = requests.get(
"https://api.westkite.com/v1/search",
params={"q": "550 Madison Ave, New York"},
headers={"Authorization": f"Bearer {os.environ['WESTKITE_API_KEY']}"},
)
data = res.json()["data"]Response
{
"data": {
"results": [
{
"id": "property:08cdd01c-64d6-8ed2-b7b8-8158ccadf8a0",
"title": "550 Madison Avenue · Manhattan · APN 6201001012910010",
"match": "address",
"reason": "exact",
"exact": true,
"city": "Manhattan",
"state": "NY",
"url": "https://app.westkite.com/properties/08cdd01c-64d6-8ed2-b7b8-8158ccadf8a0"
}
],
"places": []
},
"missing": [],
"billing": { "operation_id": "op_1a4c…", "operation": "search", "charged_usd": 0, "balance_usd": 9.99 },
"request_id": "req_6e0b…"
}
Each option says why it matched in match: address, apn, owner or name. Text with no house number that names an owner lists that owner's properties first. Places the text names come back in places; pass one's in to narrow the search.
When nothing matches exactly, you get the closest records West Kite holds with "exact": false, and corrected_query says what they were read as. You never get an empty guess.
Pass an id to properties. To filter by type, size, sale or value, use find.
Search by point
lat and lng return the parcels whose outline contains the point. Where parcels are stacked, such as condominiums, you get all of them. Where no outline contains the point, you get the nearest property within about 500 meters, marked "nearest": true with its distance_m, so you can tell a match from a guess. With nothing that close, results is empty and missing says not_covered.
curl "https://api.westkite.com/v1/search?lat=37.7880&lng=-122.3953" \
-H "Authorization: Bearer $WESTKITE_API_KEY"
{
"data": {
"results": [
{ "id": "property:8a41d0c2-…", "title": "88 Example St, San Francisco · Retail", "nearest": true, "distance_m": 42 }
]
},
"missing": [],
"billing": { "operation_id": "op_9b1e…", "operation": "search", "charged_usd": 0, "balance_usd": 9.99 },
"request_id": "req_9b1e…"
}
A text search that finds nothing returns empty results with not_found in missing.
Errors
| Code | When |
|---|---|
invalid_request |
Neither q nor both lat and lng, both at once, or a value out of range. |
rate_limited |
Too many requests this minute. Search is free but still counts. |