New York Company Registry Scraper — DOS Active Corporations

Look up any active business entity registered with the New York Department of State and export clean, normalized records — legal name, entity type, DOS ID, jurisdiction, formation date, county and the address the state mails process to.

The source is the state's own public JSON API (data.ny.gov, dataset n9v6-gdp6). No login, no cookies, no proxies, no captchas — which is why a run of a thousand companies takes a few seconds.

What you get

One row per entity. Every field is always present; null means the registry does not publish it.

Field	Example
name / legalName	BUTCHY'S WINE & SPIRITS, INC.
entityType	DOMESTIC BUSINESS CORPORATION
sourceId	4424185 (DOS ID)
jurisdiction	New York
formationDate	2013-06-27
status	ACTIVE
mailingAddress	23 SPRING STREET, OSSINING, NY 10562
registeredAgent	name + the same process address
lastEvent	Registered in Westchester County
filingHistory	initial DOS filing date
sourceUrl	link back to the source row

Output is available as JSON, CSV or Excel from the dataset tab.

Two ways to use it

1. Search by name or DOS ID — paste terms into Search terms and run. That is the whole setup.

{
  "mode": "search",
  "searchTerms": ["TESLA", "STRIPE"],
  "maxResults": 100
}

If you already know the entity number, "dosIds": ["4424185"] is faster — one request, no name matching.

2. New-registrations feed — set mode to new and put the actor on a schedule.

{
  "mode": "new",
  "sinceDays": 7,
  "maxResults": 1000
}

Each run returns only entities filed after the previous run's cursor, newest first. It remembers what it already sent, so you are never charged twice for the same company. This is the setup for lead generation and compliance monitoring.

Input	Default	Notes
mode	search	new = incremental filings feed
searchTerms	["TESLA"]	names or fragments, 2–80 characters
dosIds	[]	exact DOS entity numbers
maxResults	100	hard cap on saved records
sinceDays	7	lookback window on the first new run
sinceDate	—	explicit YYYY-MM-DD start date
maxPagesPerTerm	10	paging cap per term (1000 rows per page)
includeRaw	false	attach the untouched source JSON
maxConcurrency	3	parallel requests

What New York does not publish

The dataset covers active entities only, and it is deliberately shallow. Be aware before buying:

• No officers or directors — the registry publishes the process address, not the people.
• No EIN or document number other than the DOS ID.
• No status field — every row in this dataset is active by definition, so status is always ACTIVE and statusDate is null.
• No filing history beyond the initial filing date.
• No principal address — only the address DOS mails process to. It fills mailingAddress; repeating it in principalAddress would imply a distinction the data does not make.

If you need officers and full history, that is a different (and much more heavily protected) source; this actor does not pretend otherwise.

Broad searches

The API returns at most 1000 rows per request. A very broad term such as "A" matches hundreds of thousands of entities, so paging is capped by maxPagesPerTerm (10 pages = 10 000 rows by default). When a term hits that cap the run keeps going and the summary reports truncated: true — narrow the term, or raise the cap.

Cost

$2 per 1 000 companies, plus a one-off $0.00005 for starting the run.

You pay only for records that land in the dataset. Because the actor charges on saved rows, a search that returns nothing costs you the start fee and nothing more. In new mode a quiet week can legitimately produce zero records — and zero charge.

Check the Pricing tab of this Actor for the current price before running at scale.

Notes

• Data comes straight from data.ny.gov; sourceUrl on every record lets you verify it.
• No proxy is used, so there is no proxy cost passed on to you.
• The state updates the dataset continuously; scrapedAt tells you when this actor read it.