Claude Fable 5.1 (Adaptive Reasoning, Low Effort, Default Fallback): benchmark results

Provider: Anthropic. Released 2026-09-01. Access: API.

Unified ELO 1711 ± 1, rank #83 of 1761 rated models, from 17 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AA Omniscience - Humanities & Social Sciences60Accuracy (%)98.8
AA Omniscience - Science, Engineering & Mathematics55.2Accuracy (%)98.2
AA-Omniscience Accuracy60.23Accuracy (%)98.1
AA Omniscience - Business50.2Accuracy (%)97.9
AA Omniscience34.13Score97.7
AA Humanity's Last Exam48.89Accuracy (%)97.3
AA Omniscience - Law60.1Accuracy (%)97.3
AA Omniscience - Health51.4Accuracy (%)96.9
AA Omniscience - Software Engineering (SWE)84.5Accuracy (%)96.7
AA CritPt27.71Accuracy (%)96.6
Artificial Analysis Intelligence Index47.78Intelligence Index96
AA Long Context Reasoning82.33Accuracy (%)95.8

Interactive version: theaggregate.ai/model?slug=claude-fable-5-1-adaptive-reasoning-low-effort-default-fallback · How It Works · Data refreshed daily, snapshot 2026-09-05.