# Claude Sonnet 4.5 vs GPT-5.1

> GPT-5.1 is the stronger model overall, scoring 48.8 to 44.0 on the Noometry Index.

- Canonical page: https://noometry.com/compare/claude-sonnet-4-5-vs-gpt-5-1
- Last updated: 2026-10-10
- Shared benchmarks: 53

## Summary

- They share 53 benchmarks with published results for both. Claude Sonnet 4.5 scores higher in 3 categories and GPT-5.1 in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.1 leads 50.3 to 31.9.
- The biggest single-benchmark swing is MMLU-Pro: 86.9% for Claude Sonnet 4.5 and 57.9% for GPT-5.1.
- GPT-5.1 is cheaper at $1.25 / $10 per million input/output tokens, against $3 / $15 for Claude Sonnet 4.5.
- Claude Sonnet 4.5 accepts more context: 1M tokens versus 400K.

## Snapshot

| | Claude Sonnet 4.5 | GPT-5.1 |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Noometry Index | 44.0 | 48.8 |
| Rank | 76 | 47 |
| Context | 1M | 400K |
| Input $/M | $3 | $1.25 |
| Output $/M | $15 | $10 |
| Weights | Proprietary | Proprietary |

## Coding

- Claude Sonnet 4.5: 47.8 (#54)
- GPT-5.1: 46.9 (#59)

| Benchmark | Claude Sonnet 4.5 | GPT-5.1 |
|---|---|---|
| SWE-bench Verified | 71.3% | 68% |
| SWE-bench Verified (bash only) | 71.4% | 66% |
| LMArena WebDev | 1393 | 1395 |
| SciCode | 44.7% | 43.3% |
| GSO | 14.7% | 13.7% |
| WeirdML | 47.7% | 60.8% |
| LMArena Coding | 1489 | 1454 |
| ALE-Bench | 796.15 | 1,192 |
| SWE-bench Multilingual | 67% | — |
| LiveBench Coding | — | 72.5% |
| AlgoTune | 1.52 | — |

## Agentic & Tool Use

- Claude Sonnet 4.5: 38.1 (#30)
- GPT-5.1: 32.5 (#55)

| Benchmark | Claude Sonnet 4.5 | GPT-5.1 |
|---|---|---|
| Terminal-Bench | 46.5% | 47.6% |
| DeepResearch Bench | 52.6% | 42.8% |
| LMArena Search | 1159 | 1199 |
| Vending-Bench 2 | 3,839 | 1,473 |
| Berkeley Function Calling Leaderboard | 73.2% | — |
| GDPval | 42.5% | — |
| Remote Labor Index | 2.1% | — |
| τ²-bench Airline | 72% | — |
| τ²-bench Banking | 25.3% | — |
| τ²-bench Retail | 72.4% | — |
| τ²-bench Telecom | 84.9% | — |
| Cybench | 60% | — |
| OSWorld | 62.9% | — |
| METR Time Horizons | 67.4% | — |

## Reasoning

- Claude Sonnet 4.5: 26.9 (#113)
- GPT-5.1: 39.9 (#50)

| Benchmark | Claude Sonnet 4.5 | GPT-5.1 |
|---|---|---|
| ARC-AGI-2 | 13.6% | 17.6% |
| SimpleBench | 54.3% | 53.2% |
| ARC-AGI-1 | 63.7% | 72.8% |
| CritPt | 1.1% | 4.9% |
| Chess Puzzles | 12% | 32% |
| EnigmaEval | 6% | 11.2% |
| LMArena Hard Prompts | 1462 | 1457 |
| Mystery Game Puzzles | 17% | 19% |
| DTBench | 83.2% | 90.1% |
| LMCA | 38.8% | 43.9% |
| Epoch Capabilities Index | 146.84 | 149.66 |
| ForecastBench | 61.9 | 58.1 |
| Kagi LLM Benchmark | 57.9% | — |
| NYT Connections (extended) | 37.3% | — |
| EBR-Bench | 2.4% | — |
| LiveBench Reasoning | — | 95.8% |
| LiveBench Data Analysis | — | 72.1% |
| LiveBench | — | 78.8% |

## Math

- Claude Sonnet 4.5: 31.9 (#190)
- GPT-5.1: 50.3 (#45)

| Benchmark | Claude Sonnet 4.5 | GPT-5.1 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 77.8% | 88.6% |
| Omni-MATH | 55.3% | 46.4% |
| LMArena Math | 1448 | 1446 |
| FrontierMath (Feb 2025 set) | 15.2% | 31% |
| FrontierMath Tier 4 (v1) | 4.2% | 12.5% |
| FrontierMath (Tiers 1-3) | 23.9% | — |
| FrontierMath Tier 4 | 2.4% | — |
| ProofBench | 19% | — |
| LiveBench Math | — | 94.5% |
| MATH Level 5 | 97.7% | — |

## Knowledge

- Claude Sonnet 4.5: 48.4 (#70)
- GPT-5.1: 50.5 (#65)

| Benchmark | Claude Sonnet 4.5 | GPT-5.1 |
|---|---|---|
| GPQA Diamond | 82.3% | 87.6% |
| Humanity's Last Exam | 13.7% | 23.7% |
| SimpleQA Verified | 30.7% | 48% |
| MMLU-Pro | 86.9% | 57.9% |
| Vectara Hallucination Rate | 12% | 10.9% |
| GPQA (HELM) | 68.6% | 44.2% |
| LMArena Expert | 1483 | 1470 |

## Multimodal

- Claude Sonnet 4.5: 34.1 (#84)
- GPT-5.1: 44.0 (#17)

| Benchmark | Claude Sonnet 4.5 | GPT-5.1 |
|---|---|---|
| VPCT | 39.8% | 58.7% |
| LMArena Document | 1450 | 1403 |
| LMArena Vision | — | 1250 |

## Multilingual

- Claude Sonnet 4.5: 53.6 (#65)
- GPT-5.1: 54.0 (#51)

| Benchmark | Claude Sonnet 4.5 | GPT-5.1 |
|---|---|---|
| LMArena Non-English | 1425 | 1431 |
| LMArena Chinese | 1459 | 1495 |
| LMArena French | 1459 | 1451 |
| LMArena German | 1427 | 1439 |
| LMArena Japanese | 1390 | 1453 |
| LMArena Korean | 1403 | 1401 |
| LMArena Russian | 1436 | 1435 |
| LMArena Spanish | 1457 | 1430 |

## Instruction Following

- Claude Sonnet 4.5: 75.2 (#74)
- GPT-5.1: 84.0 (#1)

| Benchmark | Claude Sonnet 4.5 | GPT-5.1 |
|---|---|---|
| IFEval | 85% | 93.5% |
| LMArena Instruction Following | 1459 | 1442 |
| LiveBench Instruction Following | — | 93.3% |

## Long Context

- Claude Sonnet 4.5: 45.2 (#45)
- GPT-5.1: 47.6 (#14)

| Benchmark | Claude Sonnet 4.5 | GPT-5.1 |
|---|---|---|
| LMArena Longer Query | 1476 | 1448 |
| CL-bench | — | 23.7% |
| CL-bench Life | — | 17.3% |

## Writing & Preference

- Claude Sonnet 4.5: 66.7 (#33)
- GPT-5.1: 64.7 (#53)

| Benchmark | Claude Sonnet 4.5 | GPT-5.1 |
|---|---|---|
| LMArena Text | 1439 | 1443 |
| LMArena Creative Writing | 1442 | 1427 |
| WildBench | 85.4% | 86.3% |
| LMArena Multi-Turn | 1465 | 1449 |
| EQ-Bench Creative Writing | 1678 | — |
| LiveBench Language | — | 80.2% |

## FAQ

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

GPT-5.1 is the stronger model overall, scoring 48.8 to 44.0 on the Noometry Index.

### Which is cheaper, Claude Sonnet 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 Sonnet 4.5 lists at $3 and $15.

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

They score almost the same on coding (47.8 vs 46.9); test both on your own repository before choosing.

### Which has the bigger context window?

Claude Sonnet 4.5 does, with 1M tokens against 400K.

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

53 benchmarks have published results for both models. Claude Sonnet 4.5 has 73 scored results on Noometry and GPT-5.1 has 63.