Claude Opus 5 vs GPT-5.5
Claude Opus 5 
GPT-5.5 Verdict
- Claude Opus 5 leads on quality: 70.0 vs 66.6.
- Claude Opus 5 is stronger in agents & tools, coding, composite, human preference, maths, multimodal, reasoning.
- GPT-5.5 is stronger in knowledge, long context, instruction following.
- They cost about the same ($10.00 per 1M blended).
- GPT-5.5 streams 1.7× faster (88 vs 52 tokens per second).
| Metric | Claude Opus 5 | GPT-5.5 |
|---|---|---|
| BenchLeader Index | 70.0 | 66.6 |
| Agents & tools score | 71.6 | 68.4 |
| Coding score | 74.4 | 57.3 |
| Composite score | 92.9 | 77.7 |
| Human preference score | 78.2 | 67.6 |
| Knowledge score | 70.5 | 73.7 |
| Long context score | 66.2 | 68.8 |
| Maths score | 67.1 | – |
| Multimodal score | 69.4 | 64.9 |
| Reasoning score | 71.1 | 69.7 |
| Instruction following score | – | 73.3 |
| Blended price $/M | $10.00 | $11.25 |
| Output speed | 52 tok/s | 88 tok/s |
| Time to first answer | 81.6 s | 62.3 s |
| Context window | 1M | 1.1M |
| GPQA Diamond | 92.9% | – |
| OTIS Mock AIME | 97.8% | – |
| Terminal-Bench | – | 84.7% |
| SimpleBench | 80.6% | 69.0% |
| Remote Labor Index | – | 6.3% |
| WeirdML | 86.3% | – |
| APEX-Agents | – | 38.5% |
| FrontierCode | – | 43.0% |
| Epoch Capabilities Index | 162.6 | 159.1 |
| LMArena Text | – | 1477 |
| LMArena Hard Prompts | – | 1498 |
| LMArena Coding | – | 1509 |
| LMArena WebDev | – | 1458 |
| LMArena Vision | – | 1296 |
| LMArena Agent | – | 3 |
| AA Intelligence Index | 50.7 | 38.6 |
| IFBench | – | 75.8% |
| AA-LCR | 79.3% | 84.3% |
| MMMU-Pro | 84.7% | 79.9% |
| AA-Omniscience | 37.1 | 20.5 |
| Terminal-Bench Hard | – | 60.6% |
| GPQA Diamond (AA) | 93.2% | 93.5% |
| Humanity's Last Exam (AA) | 54.9% | 45.8% |
| SciCode (AA) | 56.4% | 55.8% |
| τ²-Bench Telecom (AA) | – | 93.9% |
| LiveCodeBench | 89.0% | – |
| MMLU-Pro | 91.6% | – |
| IOI | 91.7% | – |
| LegalBench | 87.0% | – |
| CorpFin | 73.2% | – |
| TaxEval | 75.1% | – |
| Terminal-Bench 2.1 (Vals) | 84.6% | – |
| SWE-bench (Vals) | 97.0% | – |
| GPQA Diamond (Vals) | 93.4% | – |
| Vals Index | 67.2 | – |
| HiL-Bench | 57.0% | 39.7% |
| EQ-Bench 4 | 1385 | 1315 |
| Kagi LLM Benchmark | – | 88.8% |
Data as of 2026-09-09. Best configuration of each model; every score links to its source on the model pages.