How-To · Developers · Data Acquisition
How to Get Indian Court Case Data Programmatically (2026)
A product needs court data, so an engineer is asked to "just pull it from the system". In India there is no single open API to pull from, so the real question is which combination of routes to use and what each one costs you in access, reliability and legal risk. This guide lays out the practical ways to get Indian court case data in code, the constraints each route carries, and how to decide between building your own pipeline and buying an interface that already solves coverage and matching. Last verified October 2026.
- No single open API: get the data through restricted NJDG keys, do-it-yourself scraping, or a third-party court-data API.
- NJDG keys are the official programmatic route but are limited to government and institutional litigants.
- Scraping can work for a narrow need but carries CAPTCHA, terms-of-use, breakage and maintenance costs and does not scale across forums.
- Matching and freshness are hard whatever route you take, so budget for entity matching and change detection, not just fetching.
01Know what you actually need first
Before picking a route, write down three things, because they decide everything that follows:
- Which forums. District and High Courts only, or also the Supreme Court, the National Company Law Tribunal, Debt Recovery Tribunals and tax tribunals? Coverage is where most routes fall short.
- Lookup or monitoring. Do you need a one-time check of a known case, or ongoing tracking of many cases with alerts when something changes?
- Search key. Do you hold the Case Number Record identifier, or only a party name? Name search is far harder and needs matching logic.
For the data model behind these choices, see what a court data API is in India. If your real job is finding cases by a company or person name, see how to search court cases by party name.
02Route 1: official portals and NJDG keys
The data originates from the courts through the eCourts programme and surfaces on the public services portal, the mobile app and the National Judicial Data Grid. For a human, that is enough. For code, the official programmatic channel is narrow:
- NJDG open APIs, introduced in 2023, are aligned with the national data sharing policy and shared with central and state government and institutional litigants under departmental IDs and access keys.
- If you qualify as a government body or institutional litigant, this is the cleanest and most authoritative route. Apply through the official channel for an access key.
- If you do not qualify, this route is closed to you, which is the situation most startups and private products are in.
See the NJDG versus eCourts difference for what each system holds.
03Route 2: do-it-yourself scraping
The next instinct is to script the public portals. It can work for a narrow, low-volume need, but go in with eyes open about the traps:
- CAPTCHA. The public case-lookup pages are protected by CAPTCHA to stop automation. Solving it with optical character recognition may conflict with the portal terms of use, so this is a legal exposure, not just a hurdle.
- Fragility. Page structure changes without notice and silently breaks scrapers, so you own ongoing maintenance.
- No service level. There is no published public rate limit or uptime guarantee, and heavy automated traffic can get you blocked.
- Coverage ceiling. Each forum has its own portal or none, so a scraper for district courts does not give you tribunals, and a per-forum build does not scale.
Scraping looks free until you count the maintenance, the breakage and the terms-of-use risk. Those costs are the real price, not the first script.
04Route 3: a third-party court-data API
The third route is to consume a third-party service that has already solved coverage, matching and refresh, and read it through a documented interface. Several commercial providers exist. When you evaluate one, ask:
- Coverage. Exactly which courts, tribunals and forums, and how far back in time?
- Freshness. How often is each forum refreshed, and how is a lag between a hearing and the record going live handled?
- Interface. Query API, bulk export, webhooks for change alerts, or all three?
- Data handling. Where does the data sit, and is your data used to train anyone's models?
To compare the market, see the best court data APIs in India for 2026 and how a litigation search API wires into your systems.
05Matching and freshness, whatever route you take
Two problems do not go away just because you have access to the data:
- Entity matching. Court records are typed by staff, so the same company or person appears under many spellings, with and without suffixes, and similar names collide. If you search by party name, you need match confidence, not a raw string compare, or you will return false hits and miss real ones.
- Freshness and change detection. A status is "as published". For monitoring you need to poll or subscribe, compare against the last known state, and surface only real changes, not re-reads.
For the matching problem specifically, see how to find court cases against a company in India and how a litigation check API works.
06How to choose
A simple rule of thumb:
- Qualify for NJDG institutional access? Use it for the authoritative feed, and consider a third party for the forums it does not cover.
- One narrow forum, low volume, in-house appetite for maintenance? A careful, terms-aware scraper may do.
- Need breadth, reliability, matching and monitoring without owning the plumbing? A third-party API or export is usually the lower total cost.
Once you pick a source, see how to integrate a case status API for the build details, and how to consolidate litigation data if you are pulling from more than one place.
07Where Claw fits
Claw is an all-in-one legaltech platform for Indian advocates, law firms, and corporate legal teams, combining AI-based case search, an AI legal assistant (Legal GPT), case management, and compliance automation across all Indian courts and tribunals.
If you do not qualify for NJDG access and do not want to own a scraping layer, Claw gives you court-data APIs, case export and case tracking across 9,102 courts, benches and forums, covering 30 crore+ judgments with Supreme Court decisions from 1950 and High Court decisions from 1990, with results in under 5 seconds, through one documented interface instead of per-forum scripts. Matching and refresh are handled for you, and Claw does not train its AI models on your case documents.
Create a FREE Claw account to try programmatic access, and use case search across Indian courts to pull cases by party across courts and tribunals.
08Sources and further reading
- eCourts services (official portal): ecourts.gov.in
- National Judicial Data Grid: njdg.ecourts.gov.in
- eCourts judgments and orders portal: judgments.ecourts.gov.in
- Claw: clawlaw.in
This is general engineering guidance, not legal advice. Scraping and data reuse raise terms-of-use and data-protection questions, so take advice before building against any source.
09Frequently asked questions
Is there a single API to get all Indian court data?
No. There is no single open public API. The National Judicial Data Grid has open APIs restricted to government and institutional litigants under access keys, the public portals are user interfaces protected by CAPTCHA, and third parties sell their own APIs built on the public data. Most teams combine routes.
Can my startup get an NJDG API key?
Probably not directly. The NJDG open APIs are aligned with the national data sharing policy and shared with central and state government and institutional litigants under departmental IDs and access keys. A private product that is not an institutional litigant generally falls outside that, which is why many use a third-party API instead.
Is scraping the eCourts portal allowed?
The pages are public, but they use CAPTCHA to stop automation, and solving it programmatically may conflict with the portal terms of use. There is also no published public rate limit, and the markup changes without notice. Treat scraping as a legal and maintenance question, not a quick win, and take advice.
Why is matching by party name so hard?
Court records are entered by court staff, so the same company or person appears under many spellings, with and without suffixes, and different parties share similar names. A raw string compare returns false hits and misses real cases, so you need match confidence scoring to trust a name-based result.
How does Claw fit in?
Claw provides court-data APIs, case export and tracking across 9,102 courts, benches and forums covering 30 crore+ judgments, with results in under 5 seconds, through one interface that handles coverage, matching and refresh. It does not train its AI models on your case documents. Confirm exact endpoints and formats from Claw developer docs.