NY Court Decisions Scraper - Official Reports & Slip Opinions

NY Court Decisions Scraper - Official Reports & Slip Opinions

Scrape New York State court decisions straight from the Law Reporting Bureau, the official publisher, across all nine court families and back to 2003. Each row carries the case title, court, county, decision date, judge, slip opinion number, docket and index numbers, and on request the full text of the opinion, the parties, counsel of record and the Bureau's own subject classification. No login, no API key. Export to CSV, JSON, Excel or XML.

New York publishes its decisions as static monthly index pages, one per court per month, with the opinions behind them. There is no search API on the public site. This Actor walks those indexes, reads the opinions, and hands back one flat row per decision, with a keyword filter that can run against the full opinion text so you can build a subject feed the site itself does not offer.

Who uses it	What they scrape New York decisions for
Litigation finance and specialty lenders	Merchant cash advance, factoring and confession of judgment rulings as they land
Law firms and legal research teams	Every decision by a named judge, or on a named subject, across the Commercial Division
Insurers and claims analysts	Trial level rulings by county on coverage, premises and no fault disputes
Legal tech and AI teams	A clean, licensed, official corpus of state court opinions with full text
Journalists and policy researchers	What the trial courts of a county actually decided, month by month

What it does

This Actor reads the New York State Law Reporting Bureau's slip opinion service on www.nycourts.gov and returns each decision as a flat row. Every decision carries:

• ⚖️ Case identity: case title, slip opinion number, decision date, the month it was posted and the batch caption it appeared under.
• 🏛️ Court: the deciding court and county as printed, plus the court family it belongs to and its key.
• 👤 Judge and numbers: the judge or panel, and the docket number where the index table publishes one.
• 🔗 Links: the decision page and the monthly index it was listed on.
• 📄 Optional blocks: full opinion text, parties split into plaintiff and defendant, counsel of record, the Bureau's digest classification, every index and indictment number printed on the decision, the footnotes, and the publication note that says whether the opinion is final or still subject to revision.
• 🗓️ A second output: the directory of every archived month the Bureau publishes for the courts you picked, so you can see what exists before you pay to read it. Pick it with outputs.

Results export to CSV, JSON, Excel or XML, or stream from the API.

The nine court families

Court	Archive reaches back to
Court of Appeals	2003
Appellate Division, First Department	2003
Appellate Division, Second Department	2003
Appellate Division, Third Department	2003
Appellate Division, Fourth Department	2003
Appellate Term, First Department	2003
Appellate Term, Second Department	2003
Supreme Court, Commercial Division	2013
Other Courts (trial level)	2003

That is 274 archived monthly indexes for most families and 157 for the Commercial Division, plus the month in progress. monthsBack decides how far back a run walks.

What you can do with New York decision data

🔎 Build a subject feed the courts do not publish.

Set keywords to ["merchant cash advance"] and keywordScope to fullText, point it at the Commercial Division and Other Courts, and you get every trial level ruling that mentions the subject, with the full opinion attached. The Bureau publishes no such feed and its own search sits on a host that forbids crawling.

👤 Follow a judge.

Filter by judges and get every decision that judge issued, in order, with the text.

📍 Read one county.

Filter by counties with "New York County" or "Kings" and get the trial court output of that county alone.

📚 Build a training or research corpus.

Set monthsBack to 274 with includeFullText, and the run walks the whole published archive of a court family, one flat row per opinion.

🗓️ Scope the job first.

Run with "outputs": ["months"] to get the directory of available months, then decide what to read.

Why choose this scraper

	What you get
The official source	The Law Reporting Bureau is the publisher of record under Judiciary Law § 431. This reads it directly, not a third party mirror with its own gaps and its own terms.
All nine families, one shape	The nine index families do not share a column layout, and the opinion pages come in two different templates. The parser is header driven and covers both, so a run that reaches back across the 2015 template change does not change shape halfway through.
Full text keyword filtering	The one thing the public site cannot do. Matching against the opinion body is what turns a court calendar into a subject feed.
Optional blocks, billed only when they return	Full text, parties, counsel, digest classification, case numbers, footnotes and the publication note are each a checkbox. Untouched blocks are neither fetched nor billed.
No proxy, no browser	Measured from the Apify container: the site serves this Actor without either. That cost saving is in the price.
The crawlable surface only	The Bureau's own case search lives on iapps.courts.state.ny.us, whose robots file is Disallow: / for every agent. This Actor never touches that host. It reads www.nycourts.gov, which permits crawling, and nothing else.

How it compares

There is no other New York Official Reports scraper on Apify. The closest actors are general case law tools that cover federal or aggregator sources.

Feature	ParseForge NY Court Decisions	pink_comic/courtlistener-legal-opinions	johnvc/google-scholar-case-law	moving_beacon-owner1/findlaw-case-law-scraper
New York state trial courts	Yes, Commercial Division and Other Courts	Appellate only	Appellate only	Appellate only
Official publisher as the source	Yes	No, aggregator	No, aggregator	No, aggregator
Full text keyword filter	Yes	No	Search only	No
Judge, county and date filters	Yes	Partial	No	No
Digest classification and counsel	Yes	No	No	No
Archived month directory	Yes	No	No	No
Price per row	$0.005	$0.002	Per search	$0.00999

What a New York decision looks like

{
  "rowType": "decision",
  "court": "Supreme Court, New York County",
  "courtFamily": "Supreme Court, Commercial Division",
  "courtKey": "commercial-division",
  "caseTitle": "A & L 444 LLC v Ankura Trust Co., LLC",
  "slipOpinionNumber": "2026 NY Slip Op 51347(U)",
  "decisionDate": "August 20, 2026",
  "judge": "Robert R. Reed, J.",
  "docketNumber": "Not Disclosed",
  "postedLabel": "Cases Posted August 31st",
  "monthLabel": "Current month",
  "decisionUrl": "https://www.nycourts.gov/reporter/current/3dseries/2026/2026_51347.shtml",
  "indexUrl": "https://www.nycourts.gov/reporter/current/index/com_div_idxtable.shtml",
  "wordCount": 1266,
  "parties": "A & L 444 LLC, 444 PARK AVENUE SOUTH ASSOCIATES LLC, DAVID MOIN MOINIAN, Plaintiff, v Ankura Trust Company, LLC, FEOH CAPITAL, LLC, ROYAL ABSTRACT OF NEW YORK LLC, 444 PARK AVE SOUTH OWNER LLC, Defendant.",
  "plaintiff": "A & L 444 LLC, 444 PARK AVENUE SOUTH ASSOCIATES LLC, DAVID MOIN MOINIAN, Plaintiff,",
  "defendant": "Ankura Trust Company, LLC, FEOH CAPITAL, LLC, ROYAL ABSTRACT OF NEW YORK LLC, 444 PARK AVE SOUTH OWNER LLC, Defendant.",
  "digestClassification": "Attorney and Client: Withdrawal from Representation: Failure to Pay Fees-Application for Stay",
  "caseNumbers": ["Index No. 653212/2023"],
  "scrapedAt": "2026-09-02T19:59:40.348Z"
}

A field the source does not publish for a given decision comes back as Not Disclosed, never as null and never missing. docketNumber above is empty because the Commercial Division index table has no Docket column; the Court of Appeals and Appellate Division tables do.

Configure the run

Input	What it does
courts	Which of the nine index families to read. Default: Commercial Division.
monthsBack	How many archived months to walk after the current one, newest first. 0 is the month in progress only; 274 is the whole archive.
maxItems	Stop after this many rows.
keywords	Keep only decisions containing these.
keywordScope	title matches the case name and costs nothing extra. fullText opens every decision in range to test it.
matchAllKeywords	Off keeps any match, on requires all of them.
counties	Substring match on the court line, for example New York County, Kings, Westchester.
judges	Substring match on the judge line.
decidedAfter, decidedBefore	ISO dates.
startUrls	Specific decision pages or monthly index pages to read instead of walking the courts.
outputs	decisions, months, or both.
includeFullText, includeParties, includeCounsel, includeDigest, includeCaseNumbers, includeFootnotes, includePublicationNote	The optional blocks. Each one opens the decision page.
proxyMode, countryCode	Off by default. The site geo-filters, so a run from outside the United States needs datacenter.

Free users

Apify free plan runs are capped at 10 rows so the plan limit is never the surprise. Paid plans run to maxItems.

Run it

{
  "courts": ["commercial-division", "other-courts"],
  "monthsBack": 6,
  "keywords": ["merchant cash advance"],
  "keywordScope": "fullText",
  "includeFullText": true,
  "includeParties": true,
  "maxItems": 50
}

Use with AI agents (MCP)

The Actor is exposed over Apify's MCP server, so an agent can call it as a tool. Point the agent at https://mcp.apify.com with this Actor enabled and it can ask for, say, every Commercial Division decision mentioning a lender in the last six months and get structured rows back.

Troubleshooting

Zero rows and a 403 in the log. The site geo-filters and reads the TLS handshake, not the headers. On the Apify platform it serves this Actor directly. Running the code outside the United States needs proxyMode: "datacenter".

A month came back empty. The archive pages list months the Bureau never posted, for example a court's summer recess. The run logs it and moves on.

The full text scan is slow. It is one page read per decision in range, paced so the site is not hammered. Narrow courts and monthsBack, or filter by counties and dates first, which happens off the index table and costs no page reads.

docketNumber is empty. Only the Court of Appeals and the four Appellate Divisions publish a Docket column in their index tables. Tick includeCaseNumbers to get every number printed on the decision itself.

FAQ

Where does the data come from? www.nycourts.gov/reporter, the New York State Law Reporting Bureau, which publishes these decisions under Judiciary Law § 431.

Does it touch the case search? No. NYSCEF and WebCivil live on iapps.courts.state.ny.us, whose robots file forbids all crawling. This Actor reads only www.nycourts.gov, which permits it.

How far back does it go? 2003 for eight of the nine families, 2013 for the Commercial Division.

Are the opinions final? Tick includePublicationNote. Slip opinions marked uncorrected are subject to revision before they appear in the printed Official Reports.

Can I get one specific decision? Yes, put its URL in startUrls.