DocketX / How to evaluate a legal AI tool
Every legal AI product says it is accurate. Here are the questions that separate a tool that checks from a tool that sounds confident, including where we say no.
We hold 615,391 Texas appellate decisions through June 2026, 122,681 Texas statute sections, 45,660 U.S. Supreme Court opinion documents, 370 county local rules and 3,967 court rule sections word for word. Unverifiable citations are removed and disclosed. Every request carries a verification report you can retrieve.
What we do not do: we do not hold federal district or circuit opinions, other states case law (Washington is in progress), treatises, or citator history. Verification proves a citation exists and that its name matches, not that it is still good law or that it supports the point. We say unverified rather than nonexistent when our library may simply not cover it, because those are different claims.
What a firm gets →Questions
Not reliably. In our benchmarking on blinded real Texas cases, grounding in a real library plus verification changed accuracy far more than moving between frontier models.
Yes. DocketRouter puts the same law layer and citation gate around whichever model you choose, including a private model on hardware you control.
They are counted from the live library, not estimated, and the benchmark uses real Texas appellate cases with the source opinion withheld from the model.
Related
DocketX products are research and explanation tools, not legal advice and not a law firm. They show you the real rule or opinion behind a statement so you can read it yourself. If you can get a lawyer or free legal aid, do.