Anthropic · model
Claude Fable 5.1 (high with fallback) benchmark scores
As of 2026-10-07, Claude Fable 5.1 (high with fallback) (Anthropic) has measured scores on 3 benchmarks and estimated scores on 7 more.
| Benchmark | Score | Source |
|---|---|---|
| Terminal-Bench 2.1 | 89.9% | measured |
| Terminal-Bench 4.0 | 47.8% | estimated ± 11.6 pp, low confidence |
| GPQA Diamond | 95.2% | estimated ± 1.3 pp, high confidence |
| CritPt | 30.2% | estimated ± 3.5 pp, high confidence |
| τ³-Bench Banking | 39.9% | estimated ± 10.4 pp, low confidence |
| AA-AnalystAgent | 46.1% | estimated ± 13.9 pp, low confidence |
| Harvey LAB | 93.0% | measured |
| Humanity's Last Exam | 55.9% | measured |
| AutomationBench | 58.1% | estimated ± 13.0 pp, low confidence |
| EnterpriseOps-Gym | 45.5% | estimated ± 4.4 pp, medium confidence |