Claude Fable 5 (High): benchmark results
Claude Fable 5 evaluated at the high reasoning-effort setting. Provider: Anthropic. Released 2026-06-09. Access: API.
Aggregate Rating 1890 ± 10, rank #41 of 2168 rated models, from 178 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| CWE-Bench | 47.8 | Pass@1 (%) | 100 |
| Chatbot Arena (Text - Expert) | 1550 | Arena Score | 100 |
| Chatbot Arena (Vision - Diagram) | 1351 | Arena Score | 100 |
| Chatbot Arena (Vision - English) | 1305 | Arena Score | 100 |
| Chatbot Arena (Vision - Humor) | 1324 | Arena Score | 100 |
| Chatbot Arena (Vision - OCR) | 1326 | Arena Score | 100 |
| Chatbot Arena (Vision) | 1308 | Arena Score | 100 |
| Deep20Bench | 12.06 | Question score: subject-mean counted questions including mod | 100 |
| EnigmaEval | 39.28 | Score (imported) | 100 |
| FutureEval | 16.1 | Unified Forecasting Score | 100 |
| Horangi 4 - BigCodeBench (100-task subset) | 68 | Pass rate (%) | 100 |
| Horangi 4 - GLP - Coding | 77.75 | Score (%) | 100 |
Interactive version: theaggregate.ai/model?slug=claude-fable-5-high · How It Works · Data refreshed daily, snapshot 2026-10-20.