Benchmarks

AA-Briefcase

Agentsunit pts · normalized over [0, 2000]

Artificial Analysis Briefcase — an Elo-rated evaluation of real professional knowledge work (documents, analysis, deliverables), reported on a ~0-2000 scale. Companion to GDPval-AA.

#ModelSourceScoreNormalized

Score distribution

4 tracked results across the normalization window

02000

Score vs. parameters

Open-weights models, log-x params

No open-weights models with disclosed parameter counts have a score here yet.