AI
Pandex Provides European Official Legal Texts to AI Agents via API and MCP
Pandex aggregates official laws and supreme court decisions from the European Union and seven European countries, offering a unified API and MCP server. Its library contains 2800000 official documents and 37000000 linked citations, and it clearly lists whether each institution’s data has complete coverage, partial coverage, or no inclusion, targeting legal AI, compliance monitoring, and legal research products.
The project calls itself “Europe’s law, in one API.” The data comes from the institutions themselves that publish laws and decisions and can be used for search and AI agents. Current figures are 2800000 official documents, complete coverage of 33 out of 41 institutions, and 37000000 citations linked to corresponding texts.
Pandex supports connecting tools such as Claude, Cursor, GitHub Copilot, Windsurf, Zed, and Gemini through MCP or API. It is positioned to let legal question-answering cite real legal provisions and let compliance teams track changes to provisions across jurisdictions, including historical versions of clauses and effective dates.
The content scale includes about 2400000 court decisions, 421000 laws and codes, and 112000000 searchable paragraphs. The EU has 101974 documents, France 907232, Germany 177101, the Netherlands 335852, the United Kingdom 99621, Austria 855697, Switzerland 137746, and Luxembourg 41891.
The project also discloses coverage limitations item by item. For example, some French Court of Cassation decisions since 2020 are published only on Judilibre and require a registered account; some German Federal Constitutional Court chamber decisions from 1998 to 2009 are available only on the court’s website; about 27700 older Dutch decisions lack texts on Rechtspraak; and about 900 UK judgments from 1996 to 2009 were not included because archive pages were inaccessible.