Finds whether a company publishes an API: developer portal, API reference, OpenAPI or Swagger files, an API catalog, MCP servers, llms.txt and the auth scheme, each with evidence.
curl -X POST "https://api.primaryapi.com/v1/run/api.discover" \
-H "Authorization: Bearer $PRIMARYAPI_API_KEY" \
-H "Content-Type: application/json" \
-d '{"domain":"stripe.com"}'{
"output": {
"found": true,
"data": {
"hasPublicApi": "<boolean>",
"domain": "<string>",
"developerPortal": {
"evidence": "<string>",
"url": "<string>"
},
"docs": [
{
"kind": "<api_reference | developer_portal |…",
"evidence": "<string>",
"url": "<string>"
}
], …
}
},
"credits": 1
}Input and output
1 field
stripe.com.10 fields
Good to know
Respects robots.txt for every host it reads and stops after 16 requests.
A domain that does not resolve is a free miss. A miss is an answer, not an error, and it costs nothing.
A site with nothing found is a hit with hasPublicApi: false.
Sorted, timestamp-free output suits monitors (fires when a spec, catalog or MCP server appears).
The same key and credits work for every API.