Model comparison
Claude Opus 4 vs GPT-5.5
GPT-5.5 is the stronger model overall, scoring 63.4 to 43.1 on the Noometry Index.
Last verified . 38 shared benchmarks.
Summary
- They share 38 benchmarks with published results for both. Claude Opus 4 scores higher in 0 categories and GPT-5.5 in 10 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-5.5 leads 72.8 to 27.3.
- The biggest single-benchmark swing is ARC-AGI-2: 8.6% for Claude Opus 4 and 85% for GPT-5.5.
- GPT-5.5 is cheaper at $5 / $30 per million input/output tokens, against $15 / $75 for Claude Opus 4.
- GPT-5.5 accepts more context: 1.05M tokens versus 200K.
Side by side
| Claude Opus 4 | GPT-5.5 | |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Noometry Index | 43.1 | 63.4 |
| Released | 2025-05-22 | 2026-04-23 |
| Weights | Proprietary | Proprietary |
| Context window | 200K | 1.05M |
| Max output | 32K | 128K |
| Input $ / M tokens | $15 | $5 |
| Output $ / M tokens | $75 | $30 |
| Results tracked | 56 | 71 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.5 leads
Claude Opus 4: 47.2 (#62), GPT-5.5: 58.2 (#17)
| Benchmark | Claude Opus 4 | GPT-5.5 |
|---|---|---|
| SWE-bench Verified | 70.7% | 80.6% |
| GSO | 6.9% | 40.2% |
| WeirdML | 43.7% | 84.9% |
| LMArena Coding | 1442 | 1494 |
| DeepSWE | — | 67% |
| FrontierCode | — | 43% |
| SWE-bench Verified (bash only) | 67.6% | — |
| Aider Polyglot | 72% | — |
| LMArena WebDev | — | 1513 |
| SciCode | — | 56.1% |
| MirrorCode | — | 10% |
| ALE-Bench | — | 1,943 |
| AlgoTune | 1.33 | — |
Agentic & Tool Use GPT-5.5 leads
Claude Opus 4: 34.8 (#42), GPT-5.5: 50.7 (#6)
| Benchmark | Claude Opus 4 | GPT-5.5 |
|---|---|---|
| DeepResearch Bench | 46.8% | 54% |
| LMArena Search | 1127 | 1242 |
| Terminal-Bench | — | 84.7% |
| APEX-Agents | — | 55.1% |
| OSWorld 2.0 | — | 13% |
| Remote Labor Index | — | 6.3% |
| τ²-bench Banking | — | 44.6% |
| Cybench | 38% | — |
| PostTrainBench | — | 27.2% |
| ExploitBench | — | 47.4% |
| GBAEval | — | 53.2% |
| GDP.pdf | — | 26% |
| METR Time Horizons | 63.9% | — |
| Vending-Bench 2 | — | 7,524 |
Reasoning GPT-5.5 leads
Claude Opus 4: 27.3 (#121), GPT-5.5: 72.8 (#11)
| Benchmark | Claude Opus 4 | GPT-5.5 |
|---|---|---|
| ARC-AGI-2 | 8.6% | 85% |
| SimpleBench | 58.8% | 69% |
| Kagi LLM Benchmark | 74.3% | 88.8% |
| ARC-AGI-1 | 35.7% | 95% |
| CritPt | 0.3% | 27.1% |
| LMArena Hard Prompts | 1399 | 1489 |
| DTBench | 81.6% | 96% |
| LMCA | 37.4% | 54.3% |
| Epoch Capabilities Index | 142.67 | 159.1 |
| ForecastBench | 61.1 | 60.6 |
| NYT Connections (extended) | — | 96.2% |
| Chess Puzzles | — | 54% |
| EnigmaEval | 5.6% | — |
| EBR-Bench | — | 34.3% |
| Mystery Game Puzzles | — | 56% |
| Surface Evolver Bench | — | 88.1% |
| Bench to the Future 3 | — | 0.14 |
Math GPT-5.5 leads
Claude Opus 4: 42.0 (#86), GPT-5.5: 81.7 (#11)
| Benchmark | Claude Opus 4 | GPT-5.5 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 64.4% | 100% |
| LMArena Math | 1390 | 1486 |
| FrontierMath (Feb 2025 set) | 4.5% | 51.7% |
| FrontierMath Tier 4 (v1) | 4.2% | 35.4% |
| FrontierMath (Tiers 1-3) | — | 85.3% |
| FrontierMath Tier 4 | — | 72.5% |
| MathArena Final-Answer Competitions | — | 94.3% |
| ProofBench | — | 50% |
| Omni-MATH | 61.6% | — |
| MATH Level 5 | 85% | — |
| FrontierMath Erdős | — | 0% |
Knowledge GPT-5.5 leads
Claude Opus 4: 44.0 (#88), GPT-5.5: 64.4 (#17)
| Benchmark | Claude Opus 4 | GPT-5.5 |
|---|---|---|
| GPQA Diamond | 76.3% | 94% |
| Vectara Hallucination Rate | 12% | 9.3% |
| LMArena Expert | 1386 | 1508 |
| Humanity's Last Exam | 10.7% | — |
| SimpleQA Verified | — | 63% |
| MMLU-Pro | 87.5% | — |
| Confabulations | 15.9% | — |
| GPQA (HELM) | 70.8% | — |
Multimodal GPT-5.5 leads
Claude Opus 4: 31.5 (#106), GPT-5.5: 46.9 (#12)
| Benchmark | Claude Opus 4 | GPT-5.5 |
|---|---|---|
| LMArena Vision | 1192 | 1297 |
| GeoBench | 49% | — |
| VPCT | 38% | — |
| Blueprint-Bench 2 | — | 36.2% |
| Furniture Assembly | — | 44.2% |
| LMArena Document | — | 1486 |
Multilingual GPT-5.5 leads
Claude Opus 4: 48.8 (#138), GPT-5.5: 56.4 (#20)
| Benchmark | Claude Opus 4 | GPT-5.5 |
|---|---|---|
| LMArena Non-English | 1362 | 1467 |
| LMArena Chinese | 1386 | 1533 |
| LMArena French | 1372 | 1486 |
| LMArena German | 1391 | 1480 |
| LMArena Japanese | 1331 | 1498 |
| LMArena Korean | 1321 | 1460 |
| LMArena Russian | 1392 | 1473 |
| LMArena Spanish | 1389 | 1468 |
Instruction Following Too close to call
Claude Opus 4: 77.1 (#28), GPT-5.5: 77.5 (#18)
| Benchmark | Claude Opus 4 | GPT-5.5 |
|---|---|---|
| LMArena Instruction Following | 1406 | 1479 |
| IFEval | 91.8% | — |
Long Context GPT-5.5 leads
Claude Opus 4: 39.6 (#172), GPT-5.5: 48.3 (#12)
| Benchmark | Claude Opus 4 | GPT-5.5 |
|---|---|---|
| LMArena Longer Query | 1422 | 1484 |
| Fiction.LiveBench | 61.1% | — |
| CL-bench Life | — | 22.2% |
Writing & Preference GPT-5.5 leads
Claude Opus 4: 61.2 (#89), GPT-5.5: 72.7 (#13)
| Benchmark | Claude Opus 4 | GPT-5.5 |
|---|---|---|
| LMArena Text | 1377 | 1472 |
| LMArena Creative Writing | 1387 | 1455 |
| EQ-Bench Creative Writing | 1580 | 1844 |
| LMArena Multi-Turn | 1396 | 1476 |
| Short-Story Creative Writing | 83.6% | — |
| WildBench | 85.2% | — |
| EQ-Bench 4 | — | 1315 |
Frequently asked questions
Is Claude Opus 4 better than GPT-5.5?
GPT-5.5 is the stronger model overall, scoring 63.4 to 43.1 on the Noometry Index.
Which is cheaper, Claude Opus 4 or GPT-5.5?
GPT-5.5 is cheaper. It lists at $5 per million input tokens and $30 per million output tokens; Claude Opus 4 lists at $15 and $75.
Is Claude Opus 4 or GPT-5.5 better for coding?
GPT-5.5 scores higher on coding benchmarks: 58.2 versus 47.2 in the Noometry coding category.
Which has the bigger context window?
GPT-5.5 does, with 1.05M tokens against 200K.
How many benchmarks do Claude Opus 4 and GPT-5.5 share?
38 benchmarks have published results for both models. Claude Opus 4 has 56 scored results on Noometry and GPT-5.5 has 71.