Search Companies

Search the company index with filters and a free-text query. Returns matching companies, a total count, facet aggregations, and a cursor for pagination.

Authentication#

Requires a token with the read scope, sent as Authorization: Bearer cornect_….

Request#

Body parameters (all optional):

Response#

{
  "items": [
    {
      "company_id": "a0383c53-8142-5e7c-8ca3-9980289a3ccc",
      "name": "Acme Inc.",
      "domain": "acme.com",
      "industry": "Software",
      "employee_count": 320,
      "country": "United States",
      "city": "San Francisco",
      "technologies": ["aws", "stripe"],
      "founded_year": 2014
    }
  ],
  "total": 4213,
  "aggregations": { "industries": [ /* facet counts */ ] },
  "next_cursor": "eyJvZmZzZXQiOjI1fQ"
}

Company size is a band, not a headcount#

employee_count does not hold a company's actual headcount. Every company carries one of nine values, each standing for a LinkedIn size band:

employee_count_min and employee_count_max are ANDed into a single contiguous range, so non-adjacent sizes cannot be expressed with them — “micro-businesses and enterprises, nothing between” has no min/max form. employee_bands takes the set directly:

{ "employee_bands": [1, 10001] }

What “relevance” sorts by#

sort_by=relevance is the default, and it is two orderings wearing one name — because what counts as relevant depends on whether you asked a question.

quality is that same completeness ordering, applied whether or not a query is present — use it when you want the most complete records for a text search rather than the best-matching ones. relevance is unchanged and keeps its meaning for existing callers.

employee_count, founded_year and name sort on the field itself. Records missing that field sort last in both directions, so sort_order=asc does not fill the first page with blanks.

NAICS: sectors vs codes#

naics_codes filters on specific 6-digit codes. naics_sectors filters on a whole sector and is what you want for sector-level questions: a 6-digit code begins with its sector’s 2-digit prefix by construction, so the sector match covers every code in it — including codes this API has never listed, and codes a future NAICS revision adds.

{ "naics_sectors": ["31-33", "21"] }

Three sectors span more than one 2-digit prefix — Manufacturing (31, 32, 33), Retail Trade (44, 45) and Transportation and Warehousing (48, 49). Send the hyphenated code: "31" is a valid NAICS prefix but not a sector, and is rejected rather than quietly searching a third of Manufacturing. An unknown sector returns 422 listing all twenty valid values.

The two filters union: sending both returns companies matching either, not the (usually empty) intersection.

A value outside the nine returns 422 naming the offending entries and listing the valid set, rather than a 200 with no results and no explanation. Counts above were measured on 2026-08-13 and will drift; the nine values are the contract.

Code samples#

curl -X POST https://api.cornect.io/api/v1/companies/search \
  -H "Authorization: Bearer cornect_your_token_here" \
  -H "Content-Type: application/json" \
  -d '{"query":"fintech","countries":["United States"],"employee_count_min":50,"exclude_industries":["Staffing and Recruiting"],"exclude_keywords":["consulting"],"connections":[{"category":"vendor","connected_domain":"salesforce.com"}]}'

Errors#

See Errors for the full table. Codes you're most likely to see here:

Interactive reference#

Related#

• Get Company
• Pagination guide
• Create Export