Claude Sonnet 5.5 scores 68.5 on the Respan Index (#6), ahead of Claude Opus 4.8 at 52.2 (#16). Claude Sonnet 5.5 leads on 19 of the 23 benchmarks both report. Claude Sonnet 5.5 is 2.5x cheaper per token at list price.
* Estimated: no published scores in that category. How the Respan Index works
| GPQA Diamond | 93.6% | 95.6% |
| LiveBench Agentic Coding (LiveBench) | 50.5% | 56.3% |
| LiveBench Coding (LiveBench) | 81.8% | 91.4% |
| LiveBench Data Analysis (LiveBench) | 66% | 78.6% |
| LiveBench Instruction Following (LiveBench) | 72% | 70.5% |
| LiveBench Language (LiveBench) | 79.7% | 83.4% |
| LiveBench Mathematics (LiveBench) | 94.3% | 96.7% |
| LiveBench Reasoning (LiveBench) | 89.2% | 91.6% |
| Mystery Game Puzzles (Epoch AI) | 36% | 65% |
| OTIS Mock AIME 2024-2025 (Epoch AI) | 98.3% | 100% |
| SAGE (Vals AI) | 54.8% | 51.8% |
| SimpleQA Verified | 53% | 46.5% |
| Terminal-Bench 4.0 (Vals AI) | 23.2% | 64.1% |
| LMArena Agent (LMArena) | 0.0664 | 0.1252 |
| LMArena Elo | 1453 | 1471 |
| LMArena Vision (LMArena) | 1286 | 1268 |
| LMArena WebDev (LMArena) | 1556 | 1786 |
| SimpleBench (SimpleBench) | 64.8% | 75.9% |
| Humanity's Last Exam (with tools) | 57.9% | 64.5% |
Higher is better. Each score links to where it was published.