Claude Sonnet 5 vs GPT-5.5
Claude Sonnet 5 
GPT-5.5 Verdict
- GPT-5.5 leads on quality: 66.7 vs 61.4.
- Claude Sonnet 5 (high) is stronger in coding.
- GPT-5.5 is stronger in agents & tools, composite, human preference, knowledge, long context, multimodal, reasoning, instruction following.
- Claude Sonnet 5 (high) is 2.8× cheaper ($4.00 vs $11.25 per 1M blended).
- GPT-5.5 streams 1.4× faster (85 vs 60 tokens per second).
| Metric | Claude Sonnet 5 (high) | GPT-5.5 |
|---|---|---|
| BenchLeader Index | 61.4 | 66.7 |
| Agents & tools score | 61.8 | 68.5 |
| Coding score | 61.6 | 57.2 |
| Composite score | 69.5 | 77.8 |
| Human preference score | 66.2 | 67.9 |
| Knowledge score | 62.4 | 73.9 |
| Long context score | 64.8 | 68.8 |
| Multimodal score | 63.1 | 64.9 |
| Reasoning score | 66.8 | 70.1 |
| Instruction following score | – | 73.6 |
| Blended price $/M | $4.00 | $11.25 |
| Output speed | 60 tok/s | 85 tok/s |
| Time to first answer | 10.0 s | 58.6 s |
| Context window | 1M | 1.1M |
| Terminal-Bench | – | 84.7% |
| SimpleBench | – | 69.0% |
| SciCode | 48.6% | – |
| Remote Labor Index | – | 6.3% |
| WeirdML | 68.8% | – |
| APEX-Agents | – | 38.5% |
| FrontierCode | – | 43.0% |
| Epoch Capabilities Index | – | 159.1 |
| LMArena Text | 1462 | 1477 |
| LMArena Hard Prompts | 1490 | 1498 |
| LMArena Coding | 1521 | 1509 |
| LMArena WebDev | 1538 | 1458 |
| LMArena Vision | 1278 | 1296 |
| LMArena Agent | 6.3 | 3 |
| AA Intelligence Index | 32.0 | 38.6 |
| IFBench | – | 75.8% |
| AA-LCR | 76.7% | 84.3% |
| MMMU-Pro | – | 79.9% |
| AA-Omniscience | -3.7 | 20.5 |
| Terminal-Bench Hard | – | 60.6% |
| GPQA Diamond (AA) | – | 93.5% |
| Humanity's Last Exam (AA) | 35.7% | 45.8% |
| SciCode (AA) | 54.3% | 55.8% |
| τ²-Bench Telecom (AA) | – | 93.9% |
| HiL-Bench | – | 39.7% |
| EQ-Bench 4 | – | 1315 |
| Kagi LLM Benchmark | – | 88.8% |
Data as of 2026-09-10. Best configuration of each model; every score links to its source on the model pages.
Claude Sonnet 5 vs GPT-5.5: questions
- Is Claude Sonnet 5 better than GPT-5.5?
- GPT-5.5 leads on quality: 66.7 vs 61.4. The BenchLeader Index combines every independent quality benchmark; GPT-5.5 is ahead overall as of 2026-09-10, but check the category scores for your use.
- Is Claude Sonnet 5 better than GPT-5.5 for coding?
- Claude Sonnet 5 scores higher in coding (62 vs 57 on the category index, where 50 is average).
- Is Claude Sonnet 5 better than GPT-5.5 for agentic tasks?
- GPT-5.5 scores higher in agentic tasks (69 vs 62 on the category index, where 50 is average).
- Which is cheaper, Claude Sonnet 5 or GPT-5.5?
- Claude Sonnet 5 is cheaper: $4.00 against $11.25 per million tokens, blended at three input tokens per output token.
- Which is faster, Claude Sonnet 5 or GPT-5.5?
- GPT-5.5 streams faster: 85 against 60 output tokens per second.
- Which has the larger context window?
- GPT-5.5 accepts more context: 1.1M against 1M tokens.