Explainer · Developers · Legal Research

Judgment Search API for India (2026)

Judgment search and case tracking are often lumped together, but for a product they are different jobs. Case tracking watches pending matters and their next hearing. Judgment search reads the full text of decided judgments and orders, so a tool can find the right precedent, pull the holding, or classify outcomes. A judgment search API is the data layer for that second job. This explainer covers what a judgment search API is, the public Indian sources and where they fall short, what good full-text search over judgments has to get right, and how teams integrate it. Last verified October 2026.

The short answer
  • What it is: an API to search the full text of decided judgments and orders, distinct from case status or tracking.
  • Public sources: the electronic Supreme Court Reports project covers reported SC judgments from 1950; the eCourts portal carries High Court and district decisions.
  • The gap for products: no single programmatic interface spans every forum with consistent fields, so breadth usually needs a third-party API.
  • What matters: recall and ranking, citation links, structure extraction, and honesty about coverage so a blank result is real.

01What a judgment search API is, and is not

A judgment search API takes a query, a few words, a legal issue, a citation, a party, or a combination, and returns matching decided judgments and orders with enough structure for a product to use. It is not the same as a case status or tracking API:

  • Judgment search works over decided matters and their text, for research, precedent and outcome analysis.
  • Case status and tracking works over live matters, for the current stage and the next hearing.

A full legal product often needs both, but they are built differently, one is a search problem over documents, the other is a monitoring problem over records. For the tracking side, see how to integrate a case status API and what a litigation monitoring API does.

02The public sources and where they fall short

India has genuine free sources for judgments, which is a good starting point and a reason to be clear about their limits:

  • The electronic Supreme Court Reports project gives free access to reported Supreme Court judgments, with coverage going back to 1950 and a search engine built with the National Informatics Centre that uses elastic search and supports free-text, case type, year, judge and bench-strength searches.
  • The eCourts judgments and orders portal carries High Court and district court judgments, searchable through its own interface.
  • Well-known free legal databases also index Indian judgments and are widely used for research.

The limits for a product are coverage boundaries between sources, search interfaces built for people rather than code, inconsistent fields, and no single feed that spans every forum. "Reported" judgments are also a subset, so not every decision is covered the same way.

The free sources are real and useful. The gap for a product is a single programmatic interface, consistent fields, and search that spans forums instead of one portal at a time.

03What good judgment search needs to get right

Full-text search over judgments is harder than a keyword match, because legal language is dense and a near miss is a wrong answer. Watch for:

  • Recall and ranking. The right precedent must appear, and appear near the top, even when the query words differ from the judgment language.
  • Citations and cross-references. Judgments cite each other, and following that graph is often the point of the search.
  • Structure extraction. Court, bench, date, parties, and where possible the holding, pulled out of free-flowing text.
  • Coverage honesty. The API should be clear about which forums and years it covers, so a blank result is a real blank, not a coverage gap you mistook for one.

Language adds its own difficulty. The same point can be phrased very differently across decades of judgments, Latin and statutory terms sit next to plain English, and a word can carry a precise legal meaning that a generic search engine flattens. A judgment search that treats the text like ordinary web pages will miss the matter that uses different words for the same idea, which is exactly the one a researcher needs.

04Fields a product usually wants back

Regardless of provider, a judgment search result is most useful when it returns:

  • Identity: court or forum, bench, case number and the decision date.
  • Parties: the names as recorded, with the matching work already done.
  • Text and document: the full text for search and a link to or copy of the source document.
  • Relevance signals: a score or snippet so your product can rank and explain the hit.

Field names and the exact response shape vary by provider, so treat the list above as the capability, not a schema.

05Integration notes

A few practical points when you wire judgment search into a product:

  • Latency. Search is interactive, so sub-second responses matter for a usable product.
  • Caching and storage. Decide what you cache locally versus query live, and respect any licensing on the text you store.
  • Confidentiality. If you send your own matter text to help find similar judgments, know whether that text is used to train anyone's models.
  • Combine with status. Many products pair judgment search with case tracking, so plan the two data layers together.

For the wider data layer, see what a court data API is in India, and to compare providers, see the best court data APIs in India for 2026.

06Where 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.

For judgment search, Claw provides court-data APIs and case export over 30 crore+ judgments spanning 9,102 courts, benches and forums, with Supreme Court decisions from 1950 and High Court decisions from 1990, and search results in under 5 seconds, so a product can run full-text judgment search across forums through one interface instead of one portal at a time. Claw does not train its AI models on your case documents, which matters if your product sends matter text to find similar judgments.

Create a FREE Claw account to try judgment search, and use case search across Indian courts to find decisions across courts and tribunals.

07Sources and further reading

This explainer is general engineering guidance, not legal advice. Reuse and storage of judgment text can carry licensing conditions, so check the terms of any source you build on.

08Frequently asked questions

How is a judgment search API different from case tracking?

Judgment search works over the full text of decided judgments and orders, for research, precedent and outcome analysis. Case tracking works over live, pending matters, for the current stage and the next hearing. A product often needs both, but they are built differently, one as a document search problem and one as a monitoring problem.

Are Indian judgments available for free?

Yes, in part. The electronic Supreme Court Reports project gives free access to reported Supreme Court judgments with coverage from 1950, and the eCourts judgments portal carries High Court and district court decisions. The limits are coverage boundaries, interfaces built for people rather than code, and no single feed across every forum.

What does good judgment search need to get right?

Strong recall and ranking so the right precedent appears near the top even when query words differ from the judgment, the ability to follow citations between judgments, extraction of structure such as court, bench, date and parties, and clear coverage so a blank result is a genuine blank and not a hidden gap.

Can I store judgment text in my own system?

Often yes, but reuse and storage of judgment text can carry licensing conditions depending on the source, and you should respect them. Decide what you cache locally versus query live, and check the terms of any source before you store its text at scale.

How does Claw help with judgment search?

Claw offers court-data APIs and case export over 30 crore+ judgments across 9,102 courts, benches and forums, with Supreme Court decisions from 1950 and High Court decisions from 1990, and results in under 5 seconds, through one interface. It does not train its AI models on your case documents. Confirm exact endpoints and fields from Claw developer docs.

Book a demo

Thirty minutes with a Claw specialist, on your matters and your courts.

Explore CLAW

The tools behind the guides

CLAW helps Indian advocates and firms manage cases, track courts and research the law.