Claude Fable 5.1 (Low): benchmark results

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

Unified ELO 1759 ± 1, rank #25 of 3078 rated models, from 13 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Conceptual Reasoning Index - Argument Evaluation (LMCA)63.46Chance-Corrected Score (0-100)96
DuelLab Overall73.8DuelLab Score95.5
Conceptual Reasoning Index72.27Chance-Corrected Score (0-100)92.9
Conceptual Reasoning Index - Consistency (ACCoRD)80.76Chance-Corrected Score (0-100)91.8
Epoch AI - Critpt27.71Score90.6
ARC-AGI-278.33Accuracy (%)87.9
CursorBench 4.045.1Score (%)86.8
Conceptual Reasoning Index - Decision Theory (DTBench)90.22Chance-Corrected Score (0-100)82.6
Bug Hunt Bench - LMS16Planted Bugs Fixed (out of 60)79.1
Bug Hunt Bench29Planted Bugs Fixed (out of 105)78.4
Epoch AI - Cursorbench66.2Score78.1
ARC-AGI-190Accuracy (%)76.1

Interactive version: theaggregate.ai/model?slug=claude-fable-5-1-low · How It Works · Data refreshed daily, snapshot 2026-09-19.