Claude Opus 5 scores 63.9 on the Respan Index (#7), ahead of GPT-5.5 at 52.5 (#15). Claude Opus 5 leads on 20 of the 26 benchmarks both report. Claude Opus 5 is 1.1x cheaper per token at list price.
* Estimated: no published scores in that category. How the Respan Index works
| 73.2% |
| 72.5% |
| FrontierMath Tiers 1-3 v2 (Epoch AI) | 85.6% | 85.3% |
| Furniture Assembly (Epoch AI) | 60.8% | 44.2% |
| GPQA Diamond | 93.9% | 90.7% |
| LiveBench Agentic Coding (LiveBench) | 65.2% | 54% |
| LiveBench Coding (LiveBench) | 81.4% | 82.1% |
| LiveBench Data Analysis (LiveBench) | 74.6% | 81.6% |
| LiveBench Instruction Following (LiveBench) | 63.8% | 70.7% |
| LiveBench Language (LiveBench) | 88.7% | 87.4% |
| LiveBench Mathematics (LiveBench) | 95.7% | 95.9% |
| LiveBench Reasoning (LiveBench) | 91.2% | 89.7% |
| Mystery Game Puzzles (Epoch AI) | 59% | 56% |
| OTIS Mock AIME 2024-2025 (Epoch AI) | 98.9% | 84.4% |
| SAGE (Vals AI) | 49.4% | 51.5% |
| SimpleQA Verified | 59.9% | 63% |
| Vending-Bench 2 (Andon Labs) | 11181.87 | 7523.84 |
| WeirdML (Håvard Tveit Ihle) | 91.8% | 84.9% |
| LMArena Agent (LMArena) | 0.0867 | 0.0415 |
| LMArena Elo | 1490 | 1466 |
| LMArena Vision (LMArena) | 1289 | 1287 |
| tau2-bench Banking Knowledge (Sierra) | 48.7% | 44.6% |
| SimpleBench (SimpleBench) | 80.6% | 69% |
Higher is better. Each score links to where it was published.