Model comparison
Claude Opus 4.6 vs GPT-5.5
GPT-5.5 is the stronger model overall, scoring 63.4 to 58.2 on the Noometry Index.
Last verified . 59 shared benchmarks.
Summary
- They share 59 benchmarks with published results for both. Claude Opus 4.6 scores higher in 4 categories and GPT-5.5 in 6 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.5 leads 81.7 to 63.0.
- The biggest single-benchmark swing is FrontierMath Tier 4: 26.8% for Claude Opus 4.6 and 72.5% for GPT-5.5.
- Claude Opus 4.6 is cheaper at $5 / $25 per million input/output tokens, against $5 / $30 for GPT-5.5.
- GPT-5.5 accepts more context: 1.05M tokens versus 1M.
Side by side
| Claude Opus 4.6 | GPT-5.5 | |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Noometry Index | 58.2 | 63.4 |
| Released | 2026-02-04 | 2026-04-23 |
| Weights | Proprietary | Proprietary |
| Context window | 1M | 1.05M |
| Max output | 128K | 128K |
| Input $ / M tokens | $5 | $5 |
| Output $ / M tokens | $25 | $30 |
| Results tracked | 68 | 71 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Claude Opus 4.6: 57.2 (#20), GPT-5.5: 58.2 (#17)
| Benchmark | Claude Opus 4.6 | GPT-5.5 |
|---|---|---|
| SWE-bench Verified | 78.7% | 80.6% |
| FrontierCode | 26.6% | 43% |
| LMArena WebDev | 1547 | 1513 |
| GSO | 41.2% | 40.2% |
| WeirdML | 78% | 84.9% |
| LMArena Coding | 1536 | 1494 |
| ALE-Bench | 996.5 | 1,943 |
| DeepSWE | — | 67% |
| SWE-bench Verified (bash only) | 75.6% | — |
| SWE-bench Multilingual | 72% | — |
| SciCode | — | 56.1% |
| MirrorCode | — | 10% |
| AlgoTune | 1.47 | — |
Agentic & Tool Use Too close to call
Claude Opus 4.6: 51.1 (#4), GPT-5.5: 50.7 (#6)
| Benchmark | Claude Opus 4.6 | GPT-5.5 |
|---|---|---|
| Terminal-Bench | 79.8% | 84.7% |
| APEX-Agents | 46.3% | 55.1% |
| Remote Labor Index | 4.2% | 6.3% |
| τ²-bench Banking | 27.3% | 44.6% |
| DeepResearch Bench | 55.3% | 54% |
| GBAEval | 44.1% | 53.2% |
| LMArena Search | 1253 | 1242 |
| Vending-Bench 2 | 8,018 | 7,524 |
| OSWorld 2.0 | — | 13% |
| Cybench | 93% | — |
| PostTrainBench | — | 27.2% |
| ExploitBench | — | 47.4% |
| GDP.pdf | — | 26% |
| METR Time Horizons | 78.9% | — |
Reasoning GPT-5.5 leads
Claude Opus 4.6: 57.8 (#23), GPT-5.5: 72.8 (#11)
| Benchmark | Claude Opus 4.6 | GPT-5.5 |
|---|---|---|
| ARC-AGI-2 | 69.2% | 85% |
| SimpleBench | 67.6% | 69% |
| Kagi LLM Benchmark | 83.6% | 88.8% |
| NYT Connections (extended) | 92.1% | 96.2% |
| ARC-AGI-1 | 94% | 95% |
| Chess Puzzles | 17% | 54% |
| EBR-Bench | 12.7% | 34.3% |
| LMArena Hard Prompts | 1527 | 1489 |
| Mystery Game Puzzles | 25% | 56% |
| DTBench | 91.2% | 96% |
| LMCA | 55.8% | 54.3% |
| Epoch Capabilities Index | 155.24 | 159.1 |
| ForecastBench | 60 | 60.6 |
| CritPt | — | 27.1% |
| EnigmaEval | 7.6% | — |
| Thematic Generalization | 80.6% | — |
| Surface Evolver Bench | — | 88.1% |
| Bench to the Future 3 | — | 0.14 |
Math GPT-5.5 leads
Claude Opus 4.6: 63.0 (#31), GPT-5.5: 81.7 (#11)
| Benchmark | Claude Opus 4.6 | GPT-5.5 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 66% | 85.3% |
| FrontierMath Tier 4 | 26.8% | 72.5% |
| MathArena Final-Answer Competitions | 78.5% | 94.3% |
| OTIS Mock AIME 2024-2025 | 94.4% | 100% |
| ProofBench | 50% | 50% |
| LMArena Math | 1519 | 1486 |
| FrontierMath (Feb 2025 set) | 40.7% | 51.7% |
| FrontierMath Tier 4 (v1) | 22.9% | 35.4% |
| FrontierMath Erdős | — | 0% |
Knowledge GPT-5.5 leads
Claude Opus 4.6: 61.9 (#26), GPT-5.5: 64.4 (#17)
| Benchmark | Claude Opus 4.6 | GPT-5.5 |
|---|---|---|
| GPQA Diamond | 90.5% | 94% |
| SimpleQA Verified | 47% | 63% |
| Vectara Hallucination Rate | 12.2% | 9.3% |
| LMArena Expert | 1546 | 1508 |
| Humanity's Last Exam | 34.4% | — |
Multimodal GPT-5.5 leads
Claude Opus 4.6: 37.3 (#74), GPT-5.5: 46.9 (#12)
| Benchmark | Claude Opus 4.6 | GPT-5.5 |
|---|---|---|
| LMArena Vision | 1316 | 1297 |
| Furniture Assembly | 28.3% | 44.2% |
| LMArena Document | 1507 | 1486 |
| Blueprint-Bench 2 | — | 36.2% |
Multilingual Claude Opus 4.6 leads
Claude Opus 4.6: 57.9 (#6), GPT-5.5: 56.4 (#20)
| Benchmark | Claude Opus 4.6 | GPT-5.5 |
|---|---|---|
| LMArena Non-English | 1489 | 1467 |
| LMArena Chinese | 1551 | 1533 |
| LMArena French | 1513 | 1486 |
| LMArena German | 1502 | 1480 |
| LMArena Japanese | 1484 | 1498 |
| LMArena Korean | 1464 | 1460 |
| LMArena Russian | 1497 | 1473 |
| LMArena Spanish | 1510 | 1468 |
Instruction Following Claude Opus 4.6 leads
Claude Opus 4.6: 79.5 (#4), GPT-5.5: 77.5 (#18)
| Benchmark | Claude Opus 4.6 | GPT-5.5 |
|---|---|---|
| LMArena Instruction Following | 1523 | 1479 |
Long Context Too close to call
Claude Opus 4.6: 48.1 (#13), GPT-5.5: 48.3 (#12)
| Benchmark | Claude Opus 4.6 | GPT-5.5 |
|---|---|---|
| CL-bench Life | 17% | 22.2% |
| LMArena Longer Query | 1520 | 1484 |
| CL-bench | 20.7% | — |
Writing & Preference Too close to call
Claude Opus 4.6: 73.5 (#10), GPT-5.5: 72.7 (#13)
| Benchmark | Claude Opus 4.6 | GPT-5.5 |
|---|---|---|
| LMArena Text | 1503 | 1472 |
| LMArena Creative Writing | 1505 | 1455 |
| EQ-Bench Creative Writing | 1809 | 1844 |
| EQ-Bench 4 | 1223 | 1315 |
| LMArena Multi-Turn | 1513 | 1476 |
Frequently asked questions
Is Claude Opus 4.6 better than GPT-5.5?
GPT-5.5 is the stronger model overall, scoring 63.4 to 58.2 on the Noometry Index.
Which is cheaper, Claude Opus 4.6 or GPT-5.5?
Claude Opus 4.6 is cheaper. It lists at $5 per million input tokens and $25 per million output tokens; GPT-5.5 lists at $5 and $30.
Is Claude Opus 4.6 or GPT-5.5 better for coding?
They score almost the same on coding (57.2 vs 58.2); test both on your own repository before choosing.
Which has the bigger context window?
GPT-5.5 does, with 1.05M tokens against 1M.
How many benchmarks do Claude Opus 4.6 and GPT-5.5 share?
59 benchmarks have published results for both models. Claude Opus 4.6 has 68 scored results on Noometry and GPT-5.5 has 71.