Claude Sonnet 5.5 scores 68.5 on the Respan Index (#6), ahead of GPT-5.5 at 52.5 (#15). Claude Sonnet 5.5 leads on 15 of the 20 benchmarks both report. Claude Sonnet 5.5 is 2.8x cheaper per token at list price.
* Estimated: no published scores in that category. How the Respan Index works
| GPQA Diamond |
| 95.6% |
| 90.7% |
| LiveBench Agentic Coding (LiveBench) | 56.3% | 54% |
| LiveBench Coding (LiveBench) | 91.4% | 82.1% |
| LiveBench Data Analysis (LiveBench) | 78.6% | 81.6% |
| LiveBench Instruction Following (LiveBench) | 70.5% | 70.7% |
| LiveBench Language (LiveBench) | 83.4% | 87.4% |
| LiveBench Mathematics (LiveBench) | 96.7% | 95.9% |
| LiveBench Reasoning (LiveBench) | 91.6% | 89.7% |
| Mystery Game Puzzles (Epoch AI) | 65% | 56% |
| OTIS Mock AIME 2024-2025 (Epoch AI) | 100% | 84.4% |
| SAGE (Vals AI) | 51.8% | 51.5% |
| SimpleQA Verified | 46.5% | 63% |
| LMArena Agent (LMArena) | 0.1252 | 0.0415 |
| LMArena Elo | 1471 | 1466 |
| LMArena Vision (LMArena) | 1268 | 1287 |
| SimpleBench (SimpleBench) | 75.9% | 69% |
Higher is better. Each score links to where it was published.