Blackstone

Why we included this project

Blackstone is a spaCy model trained on the kind of text most NLP tools struggle with: long-form legal documents like judgments and case reports. It ships with a named-entity recogniser for English common-law entities, a text categoriser, and extras that matter in legal work, such as resolving abbreviations and detecting compound case references. Because it plugs into spaCy, you can run it on your own documents without building a pipeline from scratch. It is an experimental research project from ICLR&D, the research lab of the law reporting body for England and Wales, so expect prototype-level accuracy and a focus on that jurisdiction. Even so, for teams doing legal research, document review, or case-law analytics, it is a rare open-source starting point.