Partnerships
LawDiver partners with other legal applications that need a reliable caselaw API underneath them. You build the product. We supply the law — cases with good-law status, statutes, regulations, rules, and specialty authorities — and we are flexible about how.
Wiring up a model is easy now. Standing behind the law it cites is not. A legal application is only as trustworthy as the corpus beneath it, and the options for that corpus are worse than they look from the outside.
CourtListener and the Free Law Project do valuable work, and LawDiver uses their unrestricted bulk dumps as one input among many — the full accounting is on Licenses and Permissions. But that data is a starting point, not a foundation. A product that ships it raw is making promises the data cannot keep.
This is the one that ends the conversation. A free index can tell you a case exists. It cannot tell you whether it still stands. With no stored treatment signals and no subsequent history, every result carries an unanswered question, and your users are the ones who find out the hard way.
Coverage is uneven by court, by era, and by state. Unpublished and trial-level material is thin, recent slip opinions arrive late or not at all, and the gaps are not documented anywhere your product can read at runtime.
Reporter citation metadata stopped flowing reliably into the free bulk sources for large stretches of state caselaw. Florida So. 3d cites from roughly 2019 forward are the clearest case: the opinion is there, the citation a brief would actually use is not. A cite checker built on that data returns false negatives on real cites.
Incomplete text, plus missing citations, plus no good-law status, is not a problem you can fix in the ranking layer. The index is missing documents, missing the keys people search by, and missing the one signal that decides whether a hit is usable.
The CourtListener membership API covers personal, educational, research, journalistic, and exploratory use. A revenue-generating product, a firm of more than five attorneys, or internal tools that support an organization each need a separate commercial agreement before you ship.
Their opinions stay inside their own products. A license that pipes that content into an application competing with theirs is not something they sell. Planning a product around getting one is planning on an exception nobody gets.
Not a dump. A maintained, cross-indexed stack of American primary law, harvested directly from the courts every day, with the treatment signals that decide whether a case is safe to cite.
More than 10 million federal and state opinions, harvested continuously from the courts themselves, with close to 100 million stored relationships between authorities behind the treatment signals — overruled, vacated, reversed, abrogated, questioned, distinguished, and subsequent history.
183 million in-text citations with character offsets and pin cites, parallel reporter citations on a normalized key, and a bidirectional citation graph — so what a case cites and what has cited it are both single queries.
43,000 U.S. Code sections, the Code of Federal Regulations, over 1.5 million state statutes, federal and state court rules, and U.S. and state constitutions — cross-indexed to the opinions that cite them.
approximately 200,000 specialty opinions and administrative authorities across Tax, Immigration, EPA (EAB), SEC, NLRB, International trade, PTAB / IP boards — each classified in its own taxonomy rather than flattened into general caselaw topics.
more than 1,300 United States treaties, protocols, and conventions, and more than 8,000 pattern jury instructions from published state civil and criminal sets.
Boolean and phrase search over a full-corpus Lucene index, citation lookup on a dedicated index, fuzzy case-name and party-name lookup on trigram indexes, and semantic search over two separately built vector sets. Your product makes one call; the routing is ours.
Including a product that competes with LawDiver. That is a deliberate position, and the same one stated on the API page: we provide the raw materials, you provide the interface and the users. A product that makes money, a firm of any size, and internal tools that support an organization are all permitted uses, with no separate agreement needed to cross one of those lines.
Four delivery models, all reading the same continuously updated corpus.
Search, retrieval, and cite check over HTTP, licensed for any commercial purpose — including a product that competes with LawDiver. One key, usage recorded, no per-call quota.
The same corpus as tools for Claude, ChatGPT, Cursor, and any MCP client, on the same key as REST. The right surface when your product is an agent rather than a search box.
Credentialed, read-only SQL against the Postgres stores and the search cluster, when your team would rather write its own queries than consume endpoints.
The corpus stood up inside your infrastructure, seeded from a snapshot and kept current by incremental sync — for products where the data cannot leave the network.
Pricing, rate shape, delivery method, attribution, support expectations, and contract length are all negotiable. We would rather fit your project than hand you a rate card. Tell us what the product needs and what the constraints are, and we will try to meet the conditions of your project.
Write to [email protected] if you are interested in a conversation about partnering with us. Tell us what you are building, which jurisdictions and authorities it touches, and roughly what volume you expect. A short note is enough to start.
Related: the legal research API, the MCP server, and the coverage map — which lists, court by court and year by year, exactly what is in the database and what is retrieved from the official source when cited.
Coverage figures above are measured inventory, not targets. See the full inventory for counts by court and year. Westlaw and Lexis are trademarks of their respective owners; LawDiver is not affiliated with, endorsed by, or sponsored by either. CourtListener is a service of the Free Law Project, which likewise does not endorse LawDiver. Statements about those services describe their published terms and the contents of their public bulk files as of October 2026.