# Claude Opus 4.6 vs GPT-4

> Claude Opus 4.6 is the stronger model overall, scoring 58.3 to 28.9 on the Noometry Index.

- Canonical page: https://noometry.com/compare/claude-opus-4-6-vs-gpt-4
- Last updated: 2026-10-10
- Shared benchmarks: 28

## Summary

- They share 28 benchmarks with published results for both. Claude Opus 4.6 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 4.6 leads 62.7 to 10.1.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 94.4% for Claude Opus 4.6 and 1.1% for GPT-4.
- Claude Opus 4.6 is cheaper at $5 / $25 per million input/output tokens, against $30 / $60 for GPT-4.
- Claude Opus 4.6 accepts more context: 1M tokens versus 8K.

## Snapshot

| | Claude Opus 4.6 | GPT-4 |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Noometry Index | 58.3 | 28.9 |
| Rank | 18 | 308 |
| Context | 1M | 8K |
| Input $/M | $5 | $30 |
| Output $/M | $25 | $60 |
| Weights | Proprietary | Proprietary |

## Coding

- Claude Opus 4.6: 57.7 (#15)
- GPT-4: 31.2 (#277)

| Benchmark | Claude Opus 4.6 | GPT-4 |
|---|---|---|
| WeirdML | 78% | 12.4% |
| LMArena Coding | 1535 | 1254 |
| SWE-bench Verified | 78.7% | — |
| FrontierCode | 26.6% | — |
| SWE-bench Verified (bash only) | 75.6% | — |
| LMArena WebDev | 1546 | — |
| SWE-bench Multilingual | 72% | — |
| GSO | 41.2% | — |
| BigCodeBench Instruct | — | 46% |
| BigCodeBench Complete | — | 57.2% |
| ALE-Bench | 996.5 | — |
| AlgoTune | 1.47 | — |
| HumanEval+ | — | 79.3% |

## Agentic & Tool Use

- Claude Opus 4.6: 51.0 (#4)
- GPT-4: —

| Benchmark | Claude Opus 4.6 | GPT-4 |
|---|---|---|
| METR Time Horizons | 78.9% | 36.1% |
| Terminal-Bench | 79.8% | — |
| APEX-Agents | 46.3% | — |
| Remote Labor Index | 4.2% | — |
| τ²-bench Banking | 27.3% | — |
| Cybench | 93% | — |
| DeepResearch Bench | 55.3% | — |
| GBAEval | 44.1% | — |
| LMArena Search | 1253 | — |
| Vending-Bench 2 | 8,018 | — |

## Reasoning

- Claude Opus 4.6: 57.7 (#21)
- GPT-4: 17.3 (#285)

| Benchmark | Claude Opus 4.6 | GPT-4 |
|---|---|---|
| Chess Puzzles | 17% | 4% |
| LMArena Hard Prompts | 1527 | 1241 |
| Mystery Game Puzzles | 25% | 12% |
| DTBench | 91.2% | 62.7% |
| LMCA | 55.8% | 17.1% |
| Epoch Capabilities Index | 155.36 | 125.88 |
| ForecastBench | 60 | 57.8 |
| ARC-AGI-2 | 69.2% | — |
| SimpleBench | 67.6% | — |
| Kagi LLM Benchmark | 83.6% | — |
| NYT Connections (extended) | 92.1% | — |
| ARC-AGI-1 | 94% | — |
| EnigmaEval | 7.6% | — |
| Thematic Generalization | 80.6% | — |
| EBR-Bench | 12.7% | — |
| BIG-Bench Hard | — | 75.1% |
| HellaSwag | — | 95.3% |
| WinoGrande | — | 87.5% |

## Math

- Claude Opus 4.6: 62.7 (#25)
- GPT-4: 10.1 (#304)

| Benchmark | Claude Opus 4.6 | GPT-4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 94.4% | 1.1% |
| LMArena Math | 1517 | 1269 |
| FrontierMath (Tiers 1-3) | 66% | — |
| FrontierMath Tier 4 | 26.8% | — |
| MathArena Final-Answer Competitions | 78.5% | — |
| ProofBench | 50% | — |
| MATH Level 5 | — | 23% |
| FrontierMath (Feb 2025 set) | 40.7% | — |
| FrontierMath Tier 4 (v1) | 22.9% | — |
| GSM8K | — | 92% |

## Knowledge

- Claude Opus 4.6: 61.8 (#23)
- GPT-4: 18.4 (#277)

| Benchmark | Claude Opus 4.6 | GPT-4 |
|---|---|---|
| GPQA Diamond | 90.5% | 35.7% |
| LMArena Expert | 1546 | 1211 |
| Humanity's Last Exam | 34.4% | — |
| SimpleQA Verified | 47% | — |
| Vectara Hallucination Rate | 12.2% | — |
| MMLU | — | 86.4% |
| TriviaQA | — | 84.8% |

## Multimodal

- Claude Opus 4.6: 37.0 (#67)
- GPT-4: —

| Benchmark | Claude Opus 4.6 | GPT-4 |
|---|---|---|
| LMArena Vision | 1316 | — |
| Furniture Assembly | 28.3% | — |
| LMArena Document | 1507 | — |

## Multilingual

- Claude Opus 4.6: 58.2 (#4)
- GPT-4: 40.8 (#210)

| Benchmark | Claude Opus 4.6 | GPT-4 |
|---|---|---|
| LMArena Non-English | 1490 | 1246 |
| LMArena Chinese | 1551 | 1242 |
| LMArena French | 1513 | 1282 |
| LMArena German | 1504 | 1251 |
| LMArena Japanese | 1484 | 1209 |
| LMArena Korean | 1465 | 1184 |
| LMArena Russian | 1497 | 1251 |
| LMArena Spanish | 1512 | 1261 |

## Instruction Following

- Claude Opus 4.6: 79.7 (#3)
- GPT-4: 65.5 (#217)

| Benchmark | Claude Opus 4.6 | GPT-4 |
|---|---|---|
| LMArena Instruction Following | 1523 | 1241 |

## Long Context

- Claude Opus 4.6: 48.1 (#13)
- GPT-4: 37.7 (#207)

| Benchmark | Claude Opus 4.6 | GPT-4 |
|---|---|---|
| LMArena Longer Query | 1520 | 1244 |
| CL-bench | 20.7% | — |
| CL-bench Life | 17% | — |

## Writing & Preference

- Claude Opus 4.6: 73.7 (#9)
- GPT-4: 35.0 (#263)

| Benchmark | Claude Opus 4.6 | GPT-4 |
|---|---|---|
| LMArena Text | 1504 | 1263 |
| LMArena Creative Writing | 1506 | 1244 |
| EQ-Bench Creative Writing | 1809 | 752 |
| LMArena Multi-Turn | 1514 | 1257 |
| EQ-Bench 4 | 1223 | — |

## FAQ

### Is Claude Opus 4.6 better than GPT-4?

Claude Opus 4.6 is the stronger model overall, scoring 58.3 to 28.9 on the Noometry Index.

### Which is cheaper, Claude Opus 4.6 or GPT-4?

Claude Opus 4.6 is cheaper. It lists at $5 per million input tokens and $25 per million output tokens; GPT-4 lists at $30 and $60.

### Is Claude Opus 4.6 or GPT-4 better for coding?

Claude Opus 4.6 scores higher on coding benchmarks: 57.7 versus 31.2 in the Noometry coding category.

### Which has the bigger context window?

Claude Opus 4.6 does, with 1M tokens against 8K.

### How many benchmarks do Claude Opus 4.6 and GPT-4 share?

28 benchmarks have published results for both models. Claude Opus 4.6 has 68 scored results on Noometry and GPT-4 has 38.