Benchmarks

Harvey Legal Agent Benchmark

Agentsunit % · normalized over [0, 25]

Harvey's Legal Agent Benchmark (Vals AI) — complex, multi-step legal workflows scored by practising lawyers. Scores stay low in absolute terms: the frontier is in the teens.

#ModelSourceScoreNormalized

Score distribution

7 tracked results across the normalization window

025

Score vs. parameters

Open-weights models, log-x params

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