# Claude Opus 5.5 vs Claude Sonnet 4.5

> Claude Opus 5.5 is the stronger model overall, scoring 68.6 to 44.1 on the Noometry Index.

- Canonical page: https://noometry.com/compare/claude-opus-5-5-vs-claude-sonnet-4-5
- Last updated: 2026-10-10
- Shared benchmarks: 34

## Summary

- They share 34 benchmarks with published results for both. Claude Opus 5.5 scores higher in 10 categories and Claude Sonnet 4.5 in 0 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in math, where Claude Opus 5.5 leads 91.8 to 32.3.
- The biggest single-benchmark swing is FrontierMath Tier 4: 95% for Claude Opus 5.5 and 2.4% for Claude Sonnet 4.5.
- Claude Sonnet 4.5 is cheaper at $3 / $15 per million input/output tokens, against $4 / $20 for Claude Opus 5.5.
- Claude Opus 5.5 accepts more context: 1M tokens versus 200K.

## Snapshot

| | Claude Opus 5.5 | Claude Sonnet 4.5 |
|---|---|---|
| Provider | Anthropic | Anthropic |
| Noometry Index | 68.6 | 44.1 |
| Rank | 3 | 81 |
| Context | 1M | 200K |
| Input $/M | $4 | $3 |
| Output $/M | $20 | $15 |
| Weights | Proprietary | Proprietary |

## Coding

- Claude Opus 5.5: 71.9 (#3)
- Claude Sonnet 4.5: 47.3 (#61)

| Benchmark | Claude Opus 5.5 | Claude Sonnet 4.5 |
|---|---|---|
| LMArena WebDev | 1813 | 1393 |
| SciCode | 66.9% | 44.7% |
| LMArena Coding | 1547 | 1489 |
| ALE-Bench | 2,147 | 796.15 |
| SWE-bench Verified | — | 71.3% |
| FrontierCode | 54.6% | — |
| SWE-bench Verified (bash only) | — | 71.4% |
| CursorBench | 57.8% | — |
| SWE-bench Multilingual | — | 67% |
| FrontierSWE | 62.3% | — |
| GSO | — | 14.7% |
| WeirdML | — | 47.7% |
| MirrorCode | 77.4% | — |
| AlgoTune | — | 1.52 |

## Agentic & Tool Use

- Claude Opus 5.5: 45.3 (#15)
- Claude Sonnet 4.5: 38.3 (#32)

| Benchmark | Claude Opus 5.5 | Claude Sonnet 4.5 |
|---|---|---|
| Vending-Bench 2 | 9,235 | 3,839 |
| Terminal-Bench | — | 46.5% |
| APEX-Agents | 73.5% | — |
| 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% |
| DeepResearch Bench | — | 52.6% |
| OSWorld | — | 62.9% |
| GDP.pdf | 30.6% | — |
| LMArena Search | — | 1159 |
| METR Time Horizons | — | 67.4% |

## Reasoning

- Claude Opus 5.5: 80.2 (#3)
- Claude Sonnet 4.5: 26.9 (#125)

| Benchmark | Claude Opus 5.5 | Claude Sonnet 4.5 |
|---|---|---|
| ARC-AGI-2 | 93.3% | 13.6% |
| NYT Connections (extended) | 88.5% | 37.3% |
| ARC-AGI-1 | 98.5% | 63.7% |
| CritPt | 31.7% | 1.1% |
| EBR-Bench | 71.4% | 2.4% |
| LMArena Hard Prompts | 1535 | 1462 |
| Mystery Game Puzzles | 71% | 17% |
| DTBench | 98.9% | 83.2% |
| LMCA | 68.2% | 38.8% |
| Epoch Capabilities Index | 167.33 | 146.84 |
| SimpleBench | — | 54.3% |
| Kagi LLM Benchmark | — | 57.9% |
| Chess Puzzles | — | 12% |
| EnigmaEval | — | 6% |
| ForecastBench | — | 61.9 |

## Math

- Claude Opus 5.5: 91.8 (#3)
- Claude Sonnet 4.5: 32.3 (#216)

| Benchmark | Claude Opus 5.5 | Claude Sonnet 4.5 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 91.2% | 23.9% |
| FrontierMath Tier 4 | 95% | 2.4% |
| OTIS Mock AIME 2024-2025 | 100% | 77.8% |
| ProofBench | 100% | 19% |
| LMArena Math | 1506 | 1449 |
| Omni-MATH | — | 55.3% |
| MATH Level 5 | — | 97.7% |
| FrontierMath (Feb 2025 set) | — | 15.2% |
| FrontierMath Erdős | 2.9% | — |
| FrontierMath Tier 4 (v1) | — | 4.2% |

## Knowledge

- Claude Opus 5.5: 66.4 (#10)
- Claude Sonnet 4.5: 48.4 (#76)

| Benchmark | Claude Opus 5.5 | Claude Sonnet 4.5 |
|---|---|---|
| GPQA Diamond | 90.6% | 82.3% |
| SimpleQA Verified | 72.2% | 30.7% |
| LMArena Expert | 1547 | 1482 |
| Humanity's Last Exam | — | 13.7% |
| MMLU-Pro | — | 86.9% |
| Vectara Hallucination Rate | — | 12% |
| GPQA (HELM) | — | 68.6% |

## Multimodal

- Claude Opus 5.5: 57.8 (#1)
- Claude Sonnet 4.5: 34.8 (#89)

| Benchmark | Claude Opus 5.5 | Claude Sonnet 4.5 |
|---|---|---|
| LMArena Vision | 1321 | — |
| VPCT | — | 39.8% |
| Blueprint-Bench 2 | 51.2% | — |
| Furniture Assembly | 83.3% | — |
| LMArena Document | — | 1450 |

## Multilingual

- Claude Opus 5.5: 59.1 (#2)
- Claude Sonnet 4.5: 53.4 (#69)

| Benchmark | Claude Opus 5.5 | Claude Sonnet 4.5 |
|---|---|---|
| LMArena Non-English | 1507 | 1425 |
| LMArena Chinese | 1588 | 1459 |
| LMArena French | 1514 | 1458 |
| LMArena Russian | 1520 | 1437 |
| LMArena Spanish | 1507 | 1457 |
| LMArena German | — | 1427 |
| LMArena Japanese | — | 1390 |
| LMArena Korean | — | 1403 |

## Instruction Following

- Claude Opus 5.5: 80.0 (#3)
- Claude Sonnet 4.5: 75.0 (#78)

| Benchmark | Claude Opus 5.5 | Claude Sonnet 4.5 |
|---|---|---|
| LMArena Instruction Following | 1537 | 1459 |
| IFEval | — | 85% |

## Long Context

- Claude Opus 5.5: 47.1 (#19)
- Claude Sonnet 4.5: 45.2 (#46)

| Benchmark | Claude Opus 5.5 | Claude Sonnet 4.5 |
|---|---|---|
| LMArena Longer Query | 1532 | 1476 |

## Writing & Preference

- Claude Opus 5.5: 78.2 (#3)
- Claude Sonnet 4.5: 66.5 (#34)

| Benchmark | Claude Opus 5.5 | Claude Sonnet 4.5 |
|---|---|---|
| LMArena Text | 1515 | 1439 |
| LMArena Creative Writing | 1533 | 1442 |
| EQ-Bench Creative Writing | 2050 | 1678 |
| LMArena Multi-Turn | 1499 | 1465 |
| WildBench | — | 85.4% |

## FAQ

### Is Claude Opus 5.5 better than Claude Sonnet 4.5?

Claude Opus 5.5 is the stronger model overall, scoring 68.6 to 44.1 on the Noometry Index.

### Which is cheaper, Claude Opus 5.5 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 5.5 lists at $4 and $20.

### Is Claude Opus 5.5 or Claude Sonnet 4.5 better for coding?

Claude Opus 5.5 scores higher on coding benchmarks: 71.9 versus 47.3 in the Noometry coding category.

### Which has the bigger context window?

Claude Opus 5.5 does, with 1M tokens against 200K.

### How many benchmarks do Claude Opus 5.5 and Claude Sonnet 4.5 share?

34 benchmarks have published results for both models. Claude Opus 5.5 has 44 scored results on Noometry and Claude Sonnet 4.5 has 73.