Claude Fable 5 (High) — benchmark results
Claude Fable 5 evaluated at the high reasoning-effort setting. Provider: Anthropic. Released 2026-06-09. Access: API.
Unified ELO 2104 ± 41, rank #8 of 1776 rated models, from 19 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Agent Arena - Praise vs Complaint | 23.94 | Praise vs Complaint (%) | 100 |
| Creative Writing (Lechmazur) | 3.3 | Mean Score | 100 |
| Multi-turn Debate (Lechmazur) | 1757.2 | Bradley-Terry Rating | 100 |
| WeirdML | 87.85 | Average Score | 98.5 |
| Surface Evolver Bench | 95 | Mean Score (%) | 97.7 |
| Epoch AI - ECI | 160.75 | ECI Score | 96.5 |
| PACT (Lechmazur) | 1603 | CMS Points | 96 |
| Surface Evolver Bench Pass Rate | 87.5 | Pass Rate (%) | 95.5 |
| Opus Magnum Bench | 60.15 | Human-normalized score (%) | 95.2 |
| Agent Arena - Steerability | 14.62 | Steerability (%) | 94.6 |
| Epoch AI - Cursorbench | 70.6 | Score | 92.6 |
| Agent Arena | 12.72 | Net Improvement (%) | 91.9 |
Interactive version: theaggregate.ai/model?slug=claude-fable-5-high · How the rankings work · Data refreshed daily, snapshot 2026-07-22.