docketrouter

The Unified Interface
For Legal AI

Every frontier model, grounded in real law, ranked on real cases.

Get API KeyDiscover Models
400+
Models
58+
Providers
170
HLL 0.1-TX items
62
Fundamentals items
Every model, one API

OpenAI-compatible endpoint. Any model on the catalog, live prices, one key.

Grounded in real law

Federal Rules retrieval and citation verification injected before the model answers.

Ranked on real cases

Humanity's Last Lawsuit: blinded Texas appellate cases decided after every model's training cutoff.

Capped spend per key

Each key is backed by a hard-capped upstream sub-key. Spend can never exceed the cap.

HLL 0.1-TX · liveView all →

Fundamentals suite (saturated. a floor, not a ranking)

Six deterministic, exam-style tasks. Frontier models pass these; we keep them public so every answer is auditable.

Methodology →
Evidence10 items
Hearsay Identification

Is the offered statement hearsay under FRE 801?

Leader: Claude Sonnet 4.5 · 100%
Research & Writing8 items
Bluebook Citation Format

Pick the correctly formatted Bluebook citation.

Leader: Claude Sonnet 4.5 · 100%
Procedure12 items
Federal Civil Procedure

Black-letter FRCP / FRAP deadlines and rule numbers.

Leader: Claude Sonnet 4.5 · 100%
Procedure8 items
Limitations Arithmetic

Given a rule and dates, is the claim timely?

Leader: Grok 4.3 · 100%
Contracts12 items
Contract Clause Classification

Label a clause with its CUAD-style category.

Leader: Claude Sonnet 4.5 · 100%
Reliability12 items
Citation Hallucination Resistance

Does the model vouch for cases that don't exist?

Leader: Claude Sonnet 4.5 · 100%