# Claude Opus 4.5 vs GPT-5.1

> Claude Opus 4.5 is the stronger model overall, scoring 50.5 to 49.0 on the Noometry Index. GPT-5.1 costs 2.9× less per token, which makes it the better buy when Claude Opus 4.5's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/claude-opus-4-5-vs-gpt-5-1
- Last updated: 2026-10-12
- Shared benchmarks: 47

## Summary

- They share 47 benchmarks with published results for both. Claude Opus 4.5 scores higher in 6 categories and GPT-5.1 in 4 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in agentic & tool use, where Claude Opus 4.5 leads 47.3 to 32.7.
- The biggest single-benchmark swing is ARC-AGI-2: 37.6% for Claude Opus 4.5 and 17.6% for GPT-5.1.
- GPT-5.1 is cheaper at $1.25 / $10 per million input/output tokens, against $5 / $25 for Claude Opus 4.5.
- GPT-5.1 accepts more context: 400K tokens versus 200K.

## Snapshot

| | Claude Opus 4.5 | GPT-5.1 |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Noometry Index | 50.5 | 49.0 |
| Rank | 47 | 53 |
| Context | 200K | 400K |
| Input $/M | $5 | $1.25 |
| Output $/M | $25 | $10 |
| Weights | Proprietary | Proprietary |

## Coding

- Claude Opus 4.5: 54.8 (#27)
- GPT-5.1: 46.4 (#66)

| Benchmark | Claude Opus 4.5 | GPT-5.1 |
|---|---|---|
| SWE-bench Verified | 76.7% | 68% |
| SWE-bench Verified (bash only) | 76.8% | 66% |
| LMArena WebDev | 1494 | 1395 |
| GSO | 26.5% | 13.7% |
| WeirdML | 63.7% | 60.8% |
| LMArena Coding | 1504 | 1454 |
| ALE-Bench | 1,025 | 1,192 |
| SWE-bench Multilingual | 70.7% | — |
| SciCode | — | 43.3% |
| LiveBench Coding | — | 72.5% |
| AlgoTune | 1.77 | — |

## Agentic & Tool Use

- Claude Opus 4.5: 47.3 (#12)
- GPT-5.1: 32.7 (#60)

| Benchmark | Claude Opus 4.5 | GPT-5.1 |
|---|---|---|
| Terminal-Bench | 63.1% | 47.6% |
| DeepResearch Bench | 54.8% | 42.8% |
| LMArena Search | 1180 | 1199 |
| Vending-Bench 2 | 4,967 | 1,473 |
| Berkeley Function Calling Leaderboard | 77.5% | — |
| GDPval | 45.5% | — |
| Remote Labor Index | 3.8% | — |
| τ²-bench Airline | 84% | — |
| τ²-bench Banking | 24.7% | — |
| τ²-bench Retail | 79.6% | — |
| τ²-bench Telecom | 92.3% | — |
| Cybench | 82% | — |
| OSWorld | 66.3% | — |
| BALROG | 43.5% | — |
| METR Time Horizons | 75% | — |

## Reasoning

- Claude Opus 4.5: 42.6 (#51)
- GPT-5.1: 39.8 (#58)

| Benchmark | Claude Opus 4.5 | GPT-5.1 |
|---|---|---|
| ARC-AGI-2 | 37.6% | 17.6% |
| SimpleBench | 62% | 53.2% |
| ARC-AGI-1 | 80% | 72.8% |
| Chess Puzzles | 12% | 32% |
| EnigmaEval | 11.9% | 11.2% |
| LMArena Hard Prompts | 1476 | 1457 |
| Mystery Game Puzzles | 22% | 19% |
| DTBench | 89.9% | 90.1% |
| LMCA | 44.5% | 43.9% |
| Epoch Capabilities Index | 150.09 | 149.64 |
| ForecastBench | 60.7 | 58.1 |
| Kagi LLM Benchmark | 80.2% | — |
| NYT Connections (extended) | 52.5% | — |
| CritPt | — | 4.9% |
| EBR-Bench | 14.3% | — |
| LiveBench Reasoning | — | 95.8% |
| LiveBench Data Analysis | — | 72.1% |
| LiveBench | — | 78.8% |

## Math

- Claude Opus 4.5: 38.6 (#132)
- GPT-5.1: 52.2 (#51)

| Benchmark | Claude Opus 4.5 | GPT-5.1 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 86.1% | 88.6% |
| LMArena Math | 1463 | 1447 |
| FrontierMath (Feb 2025 set) | 20.7% | 31% |
| FrontierMath Tier 4 (v1) | 4.2% | 12.5% |
| FrontierMath (Tiers 1-3) | 34.4% | — |
| FrontierMath Tier 4 | 4.9% | — |
| ProofBench | 36% | — |
| Omni-MATH | — | 46.4% |
| LiveBench Math | — | 94.5% |

## Knowledge

- Claude Opus 4.5: 56.5 (#44)
- GPT-5.1: 50.6 (#71)

| Benchmark | Claude Opus 4.5 | GPT-5.1 |
|---|---|---|
| GPQA Diamond | 86% | 87.6% |
| Humanity's Last Exam | 25.2% | 23.7% |
| SimpleQA Verified | 45.7% | 48% |
| Vectara Hallucination Rate | 10.9% | 10.9% |
| LMArena Expert | 1487 | 1470 |
| MMLU-Pro | — | 57.9% |
| GPQA (HELM) | — | 44.2% |

## Multimodal

- Claude Opus 4.5: 31.4 (#107)
- GPT-5.1: 44.8 (#19)

| Benchmark | Claude Opus 4.5 | GPT-5.1 |
|---|---|---|
| VPCT | 40% | 58.7% |
| LMArena Document | 1462 | 1403 |
| LMArena Vision | — | 1250 |
| GeoBench | 75% | — |
| Furniture Assembly | 28.3% | — |

## Multilingual

- Claude Opus 4.5: 54.3 (#47)
- GPT-5.1: 53.8 (#56)

| Benchmark | Claude Opus 4.5 | GPT-5.1 |
|---|---|---|
| LMArena Non-English | 1438 | 1431 |
| LMArena Chinese | 1470 | 1495 |
| LMArena French | 1471 | 1450 |
| LMArena German | 1449 | 1438 |
| LMArena Japanese | 1416 | 1453 |
| LMArena Korean | 1424 | 1401 |
| LMArena Russian | 1447 | 1435 |
| LMArena Spanish | 1458 | 1433 |

## Instruction Following

- Claude Opus 4.5: 77.5 (#19)
- GPT-5.1: 83.9 (#1)

| Benchmark | Claude Opus 4.5 | GPT-5.1 |
|---|---|---|
| LMArena Instruction Following | 1478 | 1443 |
| LiveBench Instruction Following | — | 93.3% |
| IFEval | — | 93.5% |

## Long Context

- Claude Opus 4.5: 46.5 (#22)
- GPT-5.1: 47.6 (#14)

| Benchmark | Claude Opus 4.5 | GPT-5.1 |
|---|---|---|
| CL-bench | 21.1% | 23.7% |
| LMArena Longer Query | 1480 | 1447 |
| CL-bench Life | — | 17.3% |

## Writing & Preference

- Claude Opus 4.5: 68.1 (#28)
- GPT-5.1: 64.5 (#55)

| Benchmark | Claude Opus 4.5 | GPT-5.1 |
|---|---|---|
| LMArena Text | 1451 | 1443 |
| LMArena Creative Writing | 1445 | 1427 |
| LMArena Multi-Turn | 1466 | 1450 |
| EQ-Bench Creative Writing | 1687 | — |
| WildBench | — | 86.3% |
| LiveBench Language | — | 80.2% |

## FAQ

### Is Claude Opus 4.5 better than GPT-5.1?

Claude Opus 4.5 is the stronger model overall, scoring 50.5 to 49.0 on the Noometry Index. GPT-5.1 costs 2.9× less per token, which makes it the better buy when Claude Opus 4.5's lead doesn't matter for your workload.

### Which is cheaper, Claude Opus 4.5 or GPT-5.1?

GPT-5.1 is cheaper. It lists at $1.25 per million input tokens and $10 per million output tokens; Claude Opus 4.5 lists at $5 and $25.

### Is Claude Opus 4.5 or GPT-5.1 better for coding?

Claude Opus 4.5 scores higher on coding benchmarks: 54.8 versus 46.4 in the Noometry coding category.

### Which has the bigger context window?

GPT-5.1 does, with 400K tokens against 200K.

### How many benchmarks do Claude Opus 4.5 and GPT-5.1 share?

47 benchmarks have published results for both models. Claude Opus 4.5 has 69 scored results on Noometry and GPT-5.1 has 63.