# Claude Opus 4.5 vs Claude Sonnet 4.5

> Claude Opus 4.5 is the stronger model overall, scoring 50.5 to 44.1 on the Noometry Index. Claude Sonnet 4.5 costs 1.7× 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-claude-sonnet-4-5
- Last updated: 2026-10-12
- Shared benchmarks: 65

## Summary

- They share 65 benchmarks with published results for both. Claude Opus 4.5 scores higher in 9 categories and Claude Sonnet 4.5 in 1 category; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Claude Opus 4.5 leads 42.6 to 26.9.
- The biggest single-benchmark swing is ARC-AGI-2: 37.6% for Claude Opus 4.5 and 13.6% for Claude Sonnet 4.5.
- Claude Sonnet 4.5 is cheaper at $3 / $15 per million input/output tokens, against $5 / $25 for Claude Opus 4.5.

## Snapshot

| | Claude Opus 4.5 | Claude Sonnet 4.5 |
|---|---|---|
| Provider | Anthropic | Anthropic |
| Noometry Index | 50.5 | 44.1 |
| Rank | 47 | 81 |
| Context | 200K | 200K |
| Input $/M | $5 | $3 |
| Output $/M | $25 | $15 |
| Weights | Proprietary | Proprietary |

## Coding

- Claude Opus 4.5: 54.8 (#27)
- Claude Sonnet 4.5: 47.3 (#61)

| Benchmark | Claude Opus 4.5 | Claude Sonnet 4.5 |
|---|---|---|
| SWE-bench Verified | 76.7% | 71.3% |
| SWE-bench Verified (bash only) | 76.8% | 71.4% |
| LMArena WebDev | 1494 | 1393 |
| SWE-bench Multilingual | 70.7% | 67% |
| GSO | 26.5% | 14.7% |
| WeirdML | 63.7% | 47.7% |
| LMArena Coding | 1504 | 1489 |
| ALE-Bench | 1,025 | 796.15 |
| AlgoTune | 1.77 | 1.52 |
| SciCode | — | 44.7% |

## Agentic & Tool Use

- Claude Opus 4.5: 47.3 (#12)
- Claude Sonnet 4.5: 38.3 (#32)

| Benchmark | Claude Opus 4.5 | Claude Sonnet 4.5 |
|---|---|---|
| Terminal-Bench | 63.1% | 46.5% |
| Berkeley Function Calling Leaderboard | 77.5% | 73.2% |
| GDPval | 45.5% | 42.5% |
| Remote Labor Index | 3.8% | 2.1% |
| τ²-bench Airline | 84% | 72% |
| τ²-bench Banking | 24.7% | 25.3% |
| τ²-bench Retail | 79.6% | 72.4% |
| τ²-bench Telecom | 92.3% | 84.9% |
| Cybench | 82% | 60% |
| DeepResearch Bench | 54.8% | 52.6% |
| OSWorld | 66.3% | 62.9% |
| LMArena Search | 1180 | 1159 |
| METR Time Horizons | 75% | 67.4% |
| Vending-Bench 2 | 4,967 | 3,839 |
| BALROG | 43.5% | — |

## Reasoning

- Claude Opus 4.5: 42.6 (#51)
- Claude Sonnet 4.5: 26.9 (#125)

| Benchmark | Claude Opus 4.5 | Claude Sonnet 4.5 |
|---|---|---|
| ARC-AGI-2 | 37.6% | 13.6% |
| SimpleBench | 62% | 54.3% |
| Kagi LLM Benchmark | 80.2% | 57.9% |
| NYT Connections (extended) | 52.5% | 37.3% |
| ARC-AGI-1 | 80% | 63.7% |
| Chess Puzzles | 12% | 12% |
| EnigmaEval | 11.9% | 6% |
| EBR-Bench | 14.3% | 2.4% |
| LMArena Hard Prompts | 1476 | 1462 |
| Mystery Game Puzzles | 22% | 17% |
| DTBench | 89.9% | 83.2% |
| LMCA | 44.5% | 38.8% |
| Epoch Capabilities Index | 150.09 | 146.84 |
| ForecastBench | 60.7 | 61.9 |
| CritPt | — | 1.1% |

## Math

- Claude Opus 4.5: 38.6 (#132)
- Claude Sonnet 4.5: 32.3 (#216)

| Benchmark | Claude Opus 4.5 | Claude Sonnet 4.5 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 34.4% | 23.9% |
| FrontierMath Tier 4 | 4.9% | 2.4% |
| OTIS Mock AIME 2024-2025 | 86.1% | 77.8% |
| ProofBench | 36% | 19% |
| LMArena Math | 1463 | 1449 |
| FrontierMath (Feb 2025 set) | 20.7% | 15.2% |
| FrontierMath Tier 4 (v1) | 4.2% | 4.2% |
| Omni-MATH | — | 55.3% |
| MATH Level 5 | — | 97.7% |

## Knowledge

- Claude Opus 4.5: 56.5 (#44)
- Claude Sonnet 4.5: 48.4 (#76)

| Benchmark | Claude Opus 4.5 | Claude Sonnet 4.5 |
|---|---|---|
| GPQA Diamond | 86% | 82.3% |
| Humanity's Last Exam | 25.2% | 13.7% |
| SimpleQA Verified | 45.7% | 30.7% |
| Vectara Hallucination Rate | 10.9% | 12% |
| LMArena Expert | 1487 | 1482 |
| MMLU-Pro | — | 86.9% |
| GPQA (HELM) | — | 68.6% |

## Multimodal

- Claude Opus 4.5: 31.4 (#107)
- Claude Sonnet 4.5: 34.8 (#89)

| Benchmark | Claude Opus 4.5 | Claude Sonnet 4.5 |
|---|---|---|
| VPCT | 40% | 39.8% |
| LMArena Document | 1462 | 1450 |
| GeoBench | 75% | — |
| Furniture Assembly | 28.3% | — |

## Multilingual

- Claude Opus 4.5: 54.3 (#47)
- Claude Sonnet 4.5: 53.4 (#69)

| Benchmark | Claude Opus 4.5 | Claude Sonnet 4.5 |
|---|---|---|
| LMArena Non-English | 1438 | 1425 |
| LMArena Chinese | 1470 | 1459 |
| LMArena French | 1471 | 1458 |
| LMArena German | 1449 | 1427 |
| LMArena Japanese | 1416 | 1390 |
| LMArena Korean | 1424 | 1403 |
| LMArena Russian | 1447 | 1437 |
| LMArena Spanish | 1458 | 1457 |

## Instruction Following

- Claude Opus 4.5: 77.5 (#19)
- Claude Sonnet 4.5: 75.0 (#78)

| Benchmark | Claude Opus 4.5 | Claude Sonnet 4.5 |
|---|---|---|
| LMArena Instruction Following | 1478 | 1459 |
| IFEval | — | 85% |

## Long Context

- Claude Opus 4.5: 46.5 (#22)
- Claude Sonnet 4.5: 45.2 (#46)

| Benchmark | Claude Opus 4.5 | Claude Sonnet 4.5 |
|---|---|---|
| LMArena Longer Query | 1480 | 1476 |
| CL-bench | 21.1% | — |

## Writing & Preference

- Claude Opus 4.5: 68.1 (#28)
- Claude Sonnet 4.5: 66.5 (#34)

| Benchmark | Claude Opus 4.5 | Claude Sonnet 4.5 |
|---|---|---|
| LMArena Text | 1451 | 1439 |
| LMArena Creative Writing | 1445 | 1442 |
| EQ-Bench Creative Writing | 1687 | 1678 |
| LMArena Multi-Turn | 1466 | 1465 |
| WildBench | — | 85.4% |

## FAQ

### Is Claude Opus 4.5 better than Claude Sonnet 4.5?

Claude Opus 4.5 is the stronger model overall, scoring 50.5 to 44.1 on the Noometry Index. Claude Sonnet 4.5 costs 1.7× 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 Claude Sonnet 4.5?

Claude Sonnet 4.5 is cheaper. It lists at $3 per million input tokens and $15 per million output tokens; Claude Opus 4.5 lists at $5 and $25.

### Is Claude Opus 4.5 or Claude Sonnet 4.5 better for coding?

Claude Opus 4.5 scores higher on coding benchmarks: 54.8 versus 47.3 in the Noometry coding category.

### Which has the bigger context window?

Both accept 200K tokens.

### How many benchmarks do Claude Opus 4.5 and Claude Sonnet 4.5 share?

65 benchmarks have published results for both models. Claude Opus 4.5 has 69 scored results on Noometry and Claude Sonnet 4.5 has 73.