# Claude Opus 5.5 vs GPT-4

> Claude Opus 5.5 is the stronger model overall, scoring 68.5 to 28.9 on the Noometry Index.

- Canonical page: https://noometry.com/compare/claude-opus-5-5-vs-gpt-4
- Last updated: 2026-10-10
- Shared benchmarks: 17

## Summary

- They share 17 benchmarks with published results for both. Claude Opus 5.5 scores higher in 8 categories and GPT-4 in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Claude Opus 5.5 leads 92.4 to 10.1.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 100% for Claude Opus 5.5 and 1.1% for GPT-4.
- Claude Opus 5.5 is cheaper at $4 / $20 per million input/output tokens, against $30 / $60 for GPT-4.
- Claude Opus 5.5 accepts more context: 1M tokens versus 8K.

## Snapshot

| | Claude Opus 5.5 | GPT-4 |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Noometry Index | 68.5 | 28.9 |
| Rank | 3 | 308 |
| Context | 1M | 8K |
| Input $/M | $4 | $30 |
| Output $/M | $20 | $60 |
| Weights | Proprietary | Proprietary |

## Coding

- Claude Opus 5.5: 71.4 (#1)
- GPT-4: 31.2 (#277)

| Benchmark | Claude Opus 5.5 | GPT-4 |
|---|---|---|
| LMArena Coding | 1545 | 1254 |
| FrontierCode | 54.4% | — |
| CursorBench | 57.8% | — |
| LMArena WebDev | 1820 | — |
| FrontierSWE | 62.3% | — |
| SciCode | 66.9% | — |
| WeirdML | — | 12.4% |
| BigCodeBench Instruct | — | 46% |
| MirrorCode | 77.4% | — |
| BigCodeBench Complete | — | 57.2% |
| HumanEval+ | — | 79.3% |

## Agentic & Tool Use

- Claude Opus 5.5: 45.3 (#13)
- GPT-4: —

| Benchmark | Claude Opus 5.5 | GPT-4 |
|---|---|---|
| APEX-Agents | 73.5% | — |
| GDP.pdf | 30.6% | — |
| METR Time Horizons | — | 36.1% |

## Reasoning

- Claude Opus 5.5: 79.5 (#2)
- GPT-4: 17.3 (#285)

| Benchmark | Claude Opus 5.5 | GPT-4 |
|---|---|---|
| LMArena Hard Prompts | 1545 | 1241 |
| Mystery Game Puzzles | 71% | 12% |
| DTBench | 98.9% | 62.7% |
| LMCA | 68.2% | 17.1% |
| ARC-AGI-2 | 92.5% | — |
| NYT Connections (extended) | 88.5% | — |
| ARC-AGI-1 | 97.5% | — |
| CritPt | 31.7% | — |
| Chess Puzzles | — | 4% |
| EBR-Bench | 71.4% | — |
| BIG-Bench Hard | — | 75.1% |
| Epoch Capabilities Index | — | 125.88 |
| ForecastBench | — | 57.8 |
| HellaSwag | — | 95.3% |
| WinoGrande | — | 87.5% |

## Math

- Claude Opus 5.5: 92.4 (#2)
- GPT-4: 10.1 (#304)

| Benchmark | Claude Opus 5.5 | GPT-4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 100% | 1.1% |
| FrontierMath (Tiers 1-3) | 91.2% | — |
| FrontierMath Tier 4 | 95% | — |
| ProofBench | 100% | — |
| LMArena Math | — | 1269 |
| MATH Level 5 | — | 23% |
| GSM8K | — | 92% |

## Knowledge

- Claude Opus 5.5: 66.6 (#9)
- GPT-4: 18.4 (#277)

| Benchmark | Claude Opus 5.5 | GPT-4 |
|---|---|---|
| GPQA Diamond | 90.6% | 35.7% |
| LMArena Expert | 1564 | 1211 |
| SimpleQA Verified | 72.2% | — |
| MMLU | — | 86.4% |
| TriviaQA | — | 84.8% |

## Multimodal

- Claude Opus 5.5: 56.5 (#2)
- GPT-4: —

| Benchmark | Claude Opus 5.5 | GPT-4 |
|---|---|---|
| Furniture Assembly | 83.3% | — |

## Multilingual

- Claude Opus 5.5: 59.8 (#1)
- GPT-4: 40.8 (#210)

| Benchmark | Claude Opus 5.5 | GPT-4 |
|---|---|---|
| LMArena Non-English | 1513 | 1246 |
| LMArena Chinese | 1572 | 1242 |
| LMArena Russian | 1532 | 1251 |
| LMArena French | — | 1282 |
| LMArena German | — | 1251 |
| LMArena Japanese | — | 1209 |
| LMArena Korean | — | 1184 |
| LMArena Spanish | — | 1261 |

## Instruction Following

- Claude Opus 5.5: 80.1 (#2)
- GPT-4: 65.5 (#217)

| Benchmark | Claude Opus 5.5 | GPT-4 |
|---|---|---|
| LMArena Instruction Following | 1535 | 1241 |

## Long Context

- Claude Opus 5.5: 47.3 (#18)
- GPT-4: 37.7 (#207)

| Benchmark | Claude Opus 5.5 | GPT-4 |
|---|---|---|
| LMArena Longer Query | 1540 | 1244 |

## Writing & Preference

- Claude Opus 5.5: 78.7 (#3)
- GPT-4: 35.0 (#263)

| Benchmark | Claude Opus 5.5 | GPT-4 |
|---|---|---|
| LMArena Text | 1518 | 1263 |
| LMArena Creative Writing | 1538 | 1244 |
| EQ-Bench Creative Writing | 2050 | 752 |
| LMArena Multi-Turn | 1520 | 1257 |

## FAQ

### Is Claude Opus 5.5 better than GPT-4?

Claude Opus 5.5 is the stronger model overall, scoring 68.5 to 28.9 on the Noometry Index.

### Which is cheaper, Claude Opus 5.5 or GPT-4?

Claude Opus 5.5 is cheaper. It lists at $4 per million input tokens and $20 per million output tokens; GPT-4 lists at $30 and $60.

### Is Claude Opus 5.5 or GPT-4 better for coding?

Claude Opus 5.5 scores higher on coding benchmarks: 71.4 versus 31.2 in the Noometry coding category.

### Which has the bigger context window?

Claude Opus 5.5 does, with 1M tokens against 8K.

### How many benchmarks do Claude Opus 5.5 and GPT-4 share?

17 benchmarks have published results for both models. Claude Opus 5.5 has 35 scored results on Noometry and GPT-4 has 38.