Anthropic · model

Claude Opus 5 (max) benchmark scores

Name: benchgap: LLM benchmark scores, measured and estimated
Creator: benchgap

As of 2026-10-07, Claude Opus 5 (max) (Anthropic) has measured scores on 2 benchmarks and estimated scores on 3 more.

Benchmark	Score	Source
τ³-Bench Banking	47.1%	estimated ± 4.5 pp, medium confidence
AA-AnalystAgent	53.8%	measured
Harvey LAB	93.5%	measured
AutomationBench	67.5%	estimated ± 11.2 pp, low confidence
EnterpriseOps-Gym	53.6%	estimated ± 1.8 pp, low confidence