Legal

Contracts, filings and discovery: long documents, scanned exhibits and strict rules about where the text goes.

A lens over 4 jobs. Draft copy, under review; the models and numbers on each job page come from today's data.

The jobs it bundles

Each links to its job page (or the nearest scoreboard view while the page doesn't exist yet).

Scanned exhibits and signed forms into structured fields.

74 models fit

Whole contracts and deposition transcripts in one pass.

238 models fit

Find the clause or precedent before a model reads it.

53 models fit

Sort filings and route matters by type.

66 models fit

Each constraint opens a job page or the scoreboard with the filter already in the URL.

Data residency

Documents can't leave your own hardware or region, so start from models you can run yourself.

Self-host only → · Fits one 24 GB GPU →
License gates

Legal review is quicker for permissive licenses with no usage threshold or non-commercial clause.

Permissive licenses only →
Audit trail

Prefer models whose behaviour we checked on a live endpoint, with the probe, the reply and the date on file.

Models with first-look receipts →

How people wire a model into this job. Notes are written by hand from the cited cards and docs; the model links go to the facts.

Model: BAAI/bge-m3
Harness: Embedding endpoint (a serverless provider or self-hosted TEI) + any OpenAI-compatible chat endpoint
Weights: FP16 embeddings; FP8 generator

Chunk to the embedding model's max sequence length (card says 8192 tokens for bge-m3).
Use the generator's structured-output flag (provider row) for JSON extraction.
Re-embed on model change; keep the embedding model id in the index metadata.