Claude Fable 5.2 (Low): benchmark results
Provider: Anthropic. Released 2026-11-09. Access: API.
Aggregate Rating 1985 ± 28, rank #10 of 2201 rated models, from 17 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Conceptual Reasoning Index - Decision Theory (DTBench) | 96.88 | Chance-Corrected Score (0-100) | 98.2 |
| Conceptual Reasoning Index | 78.26 | Chance-Corrected Score (0-100) | 98.1 |
| Conceptual Reasoning Index - Argument Evaluation (LMCA) | 69.63 | Chance-Corrected Score (0-100) | 98.1 |
| Epoch AI - Dtbench | 98.13 | Score | 97.9 |
| DuelLab GameBench 2 - Reasoning Variants | 91.81 | GameBench2 normalized score (0–100) | 97.7 |
| Conceptual Reasoning Index - Consistency (ACCoRD) | 85.54 | Chance-Corrected Score (0-100) | 96.3 |
| Bug Hunt Bench - VS Code Extension | 19.5 | Planted Bugs Fixed (out of 45) | 91.8 |
| Epoch AI - Critpt | 29.43 | Score | 91.4 |
| EnigmaForge - Fact F1 | 97.2 | World-fact recovery F1 (%) over 600 instances, all items | 90.6 |
| ARC-AGI-2 | 86.67 | Accuracy (%) | 89.1 |
| CursorBench 4.0 | 53.2 | Score (%) | 88.3 |
| ARC-AGI-3 | 52 | Accuracy (%) | 86.8 |
Interactive version: theaggregate.ai/model?slug=claude-fable-5-2-low · How It Works · Data refreshed daily, snapshot 2026-11-12.