Claude Fable 5 scores 62.2 on the Respan Index (#9), ahead of Claude Opus 4.8 at 52.2 (#16). Claude Fable 5 leads on 28 of the 33 benchmarks both report. Claude Opus 4.8 is 2.0x cheaper per token at list price.
* Estimated: no published scores in that category. How the Respan Index works
| 90.2% |
| 56.1% |
| FrontierMath Tiers 1-3 v2 (Epoch AI) | 87% | 80% |
| Furniture Assembly (Epoch AI) | 35.8% | 42.5% |
| GPQA Diamond | 85.9% | 93.6% |
| LiveBench Agentic Coding (LiveBench) | 62.2% | 50.5% |
| LiveBench Coding (LiveBench) | 86% | 81.8% |
| LiveBench Data Analysis (LiveBench) | 80.5% | 66% |
| LiveBench Instruction Following (LiveBench) | 75.8% | 72% |
| LiveBench Language (LiveBench) | 90.7% | 79.7% |
| LiveBench Mathematics (LiveBench) | 96% | 94.3% |
| LiveBench Reasoning (LiveBench) | 89.7% | 89.2% |
| Mystery Game Puzzles (Epoch AI) | 52% | 36% |
| OTIS Mock AIME 2024-2025 (Epoch AI) | 100% | 98.3% |
| SAGE (Vals AI) | 51.9% | 54.8% |
| SimpleQA Verified | 70.7% | 53% |
| Terminal-Bench 4.0 (Vals AI) | 41.4% | 23.2% |
| Vending-Bench 2 (Andon Labs) | 5680.26 | 5787.43 |
| WeirdML (Håvard Tveit Ihle) | 91.9% | 82.9% |
| LMArena Agent (LMArena) | 0.0821 | 0.0664 |
| LMArena Elo | 1504 | 1453 |
| LMArena Vision (LMArena) | 1309 | 1286 |
| LMArena WebDev (LMArena) | 1625 | 1556 |
| GSO Opt@1 (GSO) | 76.5% | 47.1% |
| tau2-bench Banking Knowledge (Sierra) | 39.7% | 39.7% |
| SimpleBench (SimpleBench) | 81.9% | 64.8% |
| OSWorld-Verified | 85% | 83.4% |
| SWE-Bench Pro | 80% | 69.2% |
| SWE-bench Verified | 95% | 88.6% |
| Terminal-Bench 2.1 | 84.3% | 74.6% |
Higher is better. Each score links to where it was published.