Anthropic · model
Claude Sonnet 5.5 (max with fallback) benchmark scores
As of 2026-10-07, Claude Sonnet 5.5 (max with fallback) (Anthropic) has measured scores on 9 benchmarks and estimated scores on 5 more.
| Benchmark | Score | Source |
|---|---|---|
| Terminal-Bench 2.1 | 88.3% | estimated ± 4.0 pp, medium confidence |
| Terminal-Bench 4.0 | 63.6% | measured |
| Terminal-Bench Science 0.1 | 53.3% | measured |
| GPQA Diamond | 95.0% | estimated ± 1.3 pp, high confidence |
| SciCode | 61.0% | measured |
| CritPt | 31.4% | measured |
| GDP.pdf | 25.8% | measured |
| τ³-Bench Banking | 48.3% | estimated ± 4.5 pp, medium confidence |
| MMMU-Pro | 84.6% | estimated ± 2.6 pp, high confidence |
| AA-AnalystAgent | 57.5% | measured |
| Harvey LAB | 93.1% | measured |
| Humanity's Last Exam | 55.0% | measured |
| AutomationBench | 71.8% | measured |
| EnterpriseOps-Gym | 55.4% | estimated ± 1.8 pp, low confidence |