# Claude Opus 4.5 vs Claude Sonnet 4

> Claude Opus 4.5 is the stronger model overall, scoring 50.5 to 40.6 on the Noometry Index. Claude Sonnet 4 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
- Last updated: 2026-10-10
- Shared benchmarks: 43

## Summary

- They share 43 benchmarks with published results for both. Claude Opus 4.5 scores higher in 9 categories and Claude Sonnet 4 in 1 category; 10 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Claude Opus 4.5 leads 42.5 to 23.0.
- The biggest single-benchmark swing is Cybench: 82% for Claude Opus 4.5 and 35% for Claude Sonnet 4.
- Claude Sonnet 4 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 |
|---|---|---|
| Provider | Anthropic | Anthropic |
| Noometry Index | 50.5 | 40.6 |
| Rank | 39 | 139 |
| Context | 200K | 200K |
| Input $/M | $5 | $3 |
| Output $/M | $25 | $15 |
| Weights | Proprietary | Proprietary |

## Coding

- Claude Opus 4.5: 55.4 (#21)
- Claude Sonnet 4: 44.0 (#70)

| Benchmark | Claude Opus 4.5 | Claude Sonnet 4 |
|---|---|---|
| SWE-bench Verified (bash only) | 76.8% | 64.9% |
| GSO | 26.5% | 4.9% |
| WeirdML | 63.7% | 46.1% |
| LMArena Coding | 1504 | 1414 |
| ALE-Bench | 1,025 | 655.35 |
| SWE-bench Verified | 76.7% | — |
| Aider Polyglot | — | 61.3% |
| LMArena WebDev | 1493 | — |
| SWE-bench Multilingual | 70.7% | — |
| SciCode | — | 40% |
| AlgoTune | 1.77 | — |

## Agentic & Tool Use

- Claude Opus 4.5: 47.2 (#11)
- Claude Sonnet 4: 38.3 (#29)

| Benchmark | Claude Opus 4.5 | Claude Sonnet 4 |
|---|---|---|
| Cybench | 82% | 35% |
| DeepResearch Bench | 54.8% | 46.6% |
| OSWorld | 66.3% | 43.9% |
| METR Time Horizons | 75% | 62% |
| Terminal-Bench | 63.1% | — |
| Berkeley Function Calling Leaderboard | 77.5% | — |
| GDPval | 45.5% | — |
| Remote Labor Index | 3.8% | — |
| TheAgentCompany | — | 33.1% |
| τ²-bench Airline | 84% | — |
| τ²-bench Banking | 24.7% | — |
| τ²-bench Retail | 79.6% | — |
| τ²-bench Telecom | 92.3% | — |
| BALROG | 43.5% | — |
| LMArena Search | 1180 | — |
| Vending-Bench 2 | 4,967 | — |

## Reasoning

- Claude Opus 4.5: 42.5 (#43)
- Claude Sonnet 4: 23.0 (#169)

| Benchmark | Claude Opus 4.5 | Claude Sonnet 4 |
|---|---|---|
| ARC-AGI-2 | 37.6% | 5.9% |
| SimpleBench | 62% | 45.5% |
| Kagi LLM Benchmark | 80.2% | 73% |
| ARC-AGI-1 | 80% | 40% |
| EnigmaEval | 11.9% | 3.1% |
| LMArena Hard Prompts | 1476 | 1372 |
| DTBench | 89.9% | 77.1% |
| LMCA | 44.5% | 29% |
| Epoch Capabilities Index | 150.1 | 141.69 |
| ForecastBench | 60.7 | 60.2 |
| NYT Connections (extended) | 52.5% | — |
| CritPt | — | 0.3% |
| Chess Puzzles | 12% | — |
| EBR-Bench | 14.3% | — |
| Mystery Game Puzzles | 22% | — |

## Math

- Claude Opus 4.5: 38.5 (#85)
- Claude Sonnet 4: 41.5 (#71)

| Benchmark | Claude Opus 4.5 | Claude Sonnet 4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 86.1% | 71.1% |
| LMArena Math | 1463 | 1375 |
| FrontierMath (Feb 2025 set) | 20.7% | 4.1% |
| FrontierMath Tier 4 (v1) | 4.2% | 0% |
| FrontierMath (Tiers 1-3) | 34.4% | — |
| FrontierMath Tier 4 | 4.9% | — |
| ProofBench | 36% | — |
| Omni-MATH | — | 60.2% |
| MATH Level 5 | — | 84.4% |

## Knowledge

- Claude Opus 4.5: 56.4 (#38)
- Claude Sonnet 4: 41.6 (#99)

| Benchmark | Claude Opus 4.5 | Claude Sonnet 4 |
|---|---|---|
| GPQA Diamond | 86% | 79.2% |
| Humanity's Last Exam | 25.2% | 7.8% |
| Vectara Hallucination Rate | 10.9% | 10.3% |
| LMArena Expert | 1488 | 1371 |
| SimpleQA Verified | 45.7% | — |
| MMLU-Pro | — | 84.3% |
| Confabulations | — | 13.2% |
| GPQA (HELM) | — | 70.6% |

## Multimodal

- Claude Opus 4.5: 31.1 (#99)
- Claude Sonnet 4: 25.6 (#116)

| Benchmark | Claude Opus 4.5 | Claude Sonnet 4 |
|---|---|---|
| GeoBench | 75% | 37% |
| VPCT | 40% | 34% |
| LMArena Vision | — | 1191 |
| Furniture Assembly | 28.3% | — |
| LMArena Document | 1462 | — |
| MindCube | — | 44.8% |

## Multilingual

- Claude Opus 4.5: 54.5 (#45)
- Claude Sonnet 4: 46.9 (#151)

| Benchmark | Claude Opus 4.5 | Claude Sonnet 4 |
|---|---|---|
| LMArena Non-English | 1438 | 1333 |
| LMArena Chinese | 1470 | 1351 |
| LMArena French | 1471 | 1364 |
| LMArena German | 1449 | 1331 |
| LMArena Japanese | 1415 | 1301 |
| LMArena Korean | 1424 | 1293 |
| LMArena Russian | 1448 | 1355 |
| LMArena Spanish | 1455 | 1356 |

## Instruction Following

- Claude Opus 4.5: 77.7 (#20)
- Claude Sonnet 4: 71.8 (#141)

| Benchmark | Claude Opus 4.5 | Claude Sonnet 4 |
|---|---|---|
| LMArena Instruction Following | 1478 | 1375 |
| IFEval | — | 84% |

## Long Context

- Claude Opus 4.5: 46.4 (#21)
- Claude Sonnet 4: 33.6 (#254)

| Benchmark | Claude Opus 4.5 | Claude Sonnet 4 |
|---|---|---|
| LMArena Longer Query | 1480 | 1398 |
| Fiction.LiveBench | — | 46.9% |
| CL-bench | 21.1% | — |

## Writing & Preference

- Claude Opus 4.5: 68.2 (#28)
- Claude Sonnet 4: 57.3 (#127)

| Benchmark | Claude Opus 4.5 | Claude Sonnet 4 |
|---|---|---|
| LMArena Text | 1451 | 1351 |
| LMArena Creative Writing | 1445 | 1345 |
| EQ-Bench Creative Writing | 1687 | 1483 |
| LMArena Multi-Turn | 1466 | 1376 |
| Short-Story Creative Writing | — | 81.4% |
| WildBench | — | 83.8% |

## FAQ

### Is Claude Opus 4.5 better than Claude Sonnet 4?

Claude Opus 4.5 is the stronger model overall, scoring 50.5 to 40.6 on the Noometry Index. Claude Sonnet 4 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?

Claude Sonnet 4 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 better for coding?

Claude Opus 4.5 scores higher on coding benchmarks: 55.4 versus 44.0 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 share?

43 benchmarks have published results for both models. Claude Opus 4.5 has 69 scored results on Noometry and Claude Sonnet 4 has 58.