# Claude Opus 4 vs GPT-5.6 Sol

> GPT-5.6 Sol is the stronger model overall, scoring 64.4 to 42.9 on the Noometry Index.

- Canonical page: https://noometry.com/compare/claude-opus-4-vs-gpt-5-6-sol
- Last updated: 2026-10-10
- Shared benchmarks: 33

## Summary

- They share 33 benchmarks with published results for both. Claude Opus 4 scores higher in 0 categories and GPT-5.6 Sol in 10 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-5.6 Sol leads 74.7 to 27.4.
- The biggest single-benchmark swing is ARC-AGI-2: 8.6% for Claude Opus 4 and 92.5% for GPT-5.6 Sol.
- GPT-5.6 Sol is cheaper at $4 / $20 per million input/output tokens, against $15 / $75 for Claude Opus 4.
- GPT-5.6 Sol accepts more context: 1.05M tokens versus 200K.

## Snapshot

| | Claude Opus 4 | GPT-5.6 Sol |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Noometry Index | 42.9 | 64.4 |
| Rank | 93 | 6 |
| Context | 200K | 1.05M |
| Input $/M | $15 | $4 |
| Output $/M | $75 | $20 |
| Weights | Proprietary | Proprietary |

## Coding

- Claude Opus 4: 47.8 (#55)
- GPT-5.6 Sol: 61.1 (#7)

| Benchmark | Claude Opus 4 | GPT-5.6 Sol |
|---|---|---|
| WeirdML | 43.7% | 89.4% |
| LMArena Coding | 1442 | 1493 |
| SWE-bench Verified | 70.7% | — |
| DeepSWE | — | 72.7% |
| FrontierCode | — | 47.5% |
| SWE-bench Verified (bash only) | 67.6% | — |
| Aider Polyglot | 72% | — |
| CursorBench | — | 41.7% |
| LMArena WebDev | — | 1619 |
| FrontierSWE | — | 32.2% |
| SciCode | — | 57.1% |
| GSO | 6.9% | — |
| MirrorCode | — | 20% |
| ALE-Bench | — | 2,177 |
| AlgoTune | 1.33 | — |

## Agentic & Tool Use

- Claude Opus 4: 34.5 (#39)
- GPT-5.6 Sol: 50.2 (#7)

| Benchmark | Claude Opus 4 | GPT-5.6 Sol |
|---|---|---|
| LMArena Search | 1127 | 1257 |
| APEX-Agents | — | 51.4% |
| OSWorld 2.0 | — | 27.3% |
| τ²-bench Banking | — | 46.9% |
| Cybench | 38% | — |
| DeepResearch Bench | 46.8% | — |
| PostTrainBench | — | 36.2% |
| BALROG | — | 60% |
| GBAEval | — | 52.6% |
| GDP.pdf | — | 30.7% |
| METR Time Horizons | 63.9% | — |
| Vending-Bench 2 | — | 9,619 |

## Reasoning

- Claude Opus 4: 27.4 (#104)
- GPT-5.6 Sol: 74.7 (#6)

| Benchmark | Claude Opus 4 | GPT-5.6 Sol |
|---|---|---|
| ARC-AGI-2 | 8.6% | 92.5% |
| SimpleBench | 58.8% | 71.7% |
| Kagi LLM Benchmark | 74.3% | 67% |
| ARC-AGI-1 | 35.7% | 97.5% |
| CritPt | 0.3% | 32.3% |
| EnigmaEval | 5.6% | 37.1% |
| LMArena Hard Prompts | 1399 | 1481 |
| DTBench | 81.6% | 96% |
| LMCA | 37.4% | 59.2% |
| Epoch Capabilities Index | 142.67 | 161.99 |
| NYT Connections (extended) | — | 93.8% |
| Chess Puzzles | — | 64% |
| EBR-Bench | — | 44.8% |
| Mystery Game Puzzles | — | 58% |
| Surface Evolver Bench | — | 93.1% |
| Bench to the Future 3 | — | 0.14 |
| ForecastBench | 61.1 | — |

## Math

- Claude Opus 4: 40.2 (#80)
- GPT-5.6 Sol: 85.5 (#7)

| Benchmark | Claude Opus 4 | GPT-5.6 Sol |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 64.4% | 100% |
| LMArena Math | 1391 | 1472 |
| FrontierMath (Tiers 1-3) | — | 89.1% |
| FrontierMath Tier 4 | — | 82.9% |
| ProofBench | — | 83% |
| Omni-MATH | 61.6% | — |
| MATH Level 5 | 85% | — |
| FrontierMath (Feb 2025 set) | 4.5% | — |
| FrontierMath Erdős | — | 0% |
| FrontierMath Tier 4 (v1) | 4.2% | — |

## Knowledge

- Claude Opus 4: 43.9 (#82)
- GPT-5.6 Sol: 64.3 (#16)

| Benchmark | Claude Opus 4 | GPT-5.6 Sol |
|---|---|---|
| GPQA Diamond | 76.3% | 93.5% |
| Vectara Hallucination Rate | 12% | 12.4% |
| LMArena Expert | 1384 | 1517 |
| Humanity's Last Exam | 10.7% | — |
| SimpleQA Verified | — | 69.7% |
| MMLU-Pro | 87.5% | — |
| Confabulations | 15.9% | — |
| GPQA (HELM) | 70.8% | — |

## Multimodal

- Claude Opus 4: 30.9 (#101)
- GPT-5.6 Sol: 48.9 (#8)

| Benchmark | Claude Opus 4 | GPT-5.6 Sol |
|---|---|---|
| LMArena Vision | 1192 | 1281 |
| GeoBench | 49% | — |
| VPCT | 38% | — |
| Blueprint-Bench 2 | — | 33.6% |
| Furniture Assembly | — | 56.7% |
| LMArena Document | — | 1483 |

## Multilingual

- Claude Opus 4: 49.0 (#133)
- GPT-5.6 Sol: 55.4 (#32)

| Benchmark | Claude Opus 4 | GPT-5.6 Sol |
|---|---|---|
| LMArena Non-English | 1362 | 1450 |
| LMArena Chinese | 1386 | 1523 |
| LMArena French | 1370 | 1475 |
| LMArena German | 1391 | 1476 |
| LMArena Japanese | 1332 | 1475 |
| LMArena Korean | 1318 | 1434 |
| LMArena Russian | 1392 | 1466 |
| LMArena Spanish | 1388 | 1449 |

## Instruction Following

- Claude Opus 4: 77.3 (#26)
- GPT-5.6 Sol: 77.8 (#16)

| Benchmark | Claude Opus 4 | GPT-5.6 Sol |
|---|---|---|
| LMArena Instruction Following | 1406 | 1480 |
| IFEval | 91.8% | — |

## Long Context

- Claude Opus 4: 39.5 (#167)
- GPT-5.6 Sol: 45.3 (#42)

| Benchmark | Claude Opus 4 | GPT-5.6 Sol |
|---|---|---|
| LMArena Longer Query | 1422 | 1479 |
| Fiction.LiveBench | 61.1% | — |

## Writing & Preference

- Claude Opus 4: 61.4 (#85)
- GPT-5.6 Sol: 73.5 (#12)

| Benchmark | Claude Opus 4 | GPT-5.6 Sol |
|---|---|---|
| LMArena Text | 1377 | 1456 |
| LMArena Creative Writing | 1386 | 1450 |
| EQ-Bench Creative Writing | 1580 | 1972 |
| LMArena Multi-Turn | 1396 | 1456 |
| Short-Story Creative Writing | 83.6% | — |
| WildBench | 85.2% | — |
| EQ-Bench 4 | — | 1250 |

## FAQ

### Is Claude Opus 4 better than GPT-5.6 Sol?

GPT-5.6 Sol is the stronger model overall, scoring 64.4 to 42.9 on the Noometry Index.

### Which is cheaper, Claude Opus 4 or GPT-5.6 Sol?

GPT-5.6 Sol is cheaper. It lists at $4 per million input tokens and $20 per million output tokens; Claude Opus 4 lists at $15 and $75.

### Is Claude Opus 4 or GPT-5.6 Sol better for coding?

GPT-5.6 Sol scores higher on coding benchmarks: 61.1 versus 47.8 in the Noometry coding category.

### Which has the bigger context window?

GPT-5.6 Sol does, with 1.05M tokens against 200K.

### How many benchmarks do Claude Opus 4 and GPT-5.6 Sol share?

33 benchmarks have published results for both models. Claude Opus 4 has 56 scored results on Noometry and GPT-5.6 Sol has 64.