Skip to content

Legal Agent Benchmark mean criterion-pass rate — Harvey held-out set

Harvey AI's mean criterion-level score on its held-out legal-agent evaluation.

3 models

Higher is better.

  1. 1Claude Opus 5Anthropic94.1
  2. 2Claude Sonnet 5.5Anthropic93.1
  3. 3Claude Opus 5.5Anthropic91.2
Higher is better. Bar colour: blue proprietary, orange open weights, green other.

About this benchmark

Metric
Mean criterion-pass rate
Tasks
Harvey-held-out legal-agent tasks
Authors
Harvey AI (2026)

Scores are as reported by model developers or evaluators and may use different settings (for example reasoning effort or tools). Compare with care.