# Claude Sonnet 4 vs Qwen2-72B

> Claude Sonnet 4 is the stronger model overall, scoring 40.6 to 29.5 on the Noometry Index.

- Canonical page: https://noometry.com/compare/claude-sonnet-4-vs-qwen2-72b
- Last updated: 2026-10-10
- Shared benchmarks: 23

## Summary

- They share 23 benchmarks with published results for both. Claude Sonnet 4 scores higher in 8 categories and Qwen2-72B in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in agentic & tool use, where Claude Sonnet 4 leads 38.3 to 16.9.
- The biggest single-benchmark swing is MATH Level 5: 84.4% for Claude Sonnet 4 and 39.1% for Qwen2-72B.
- Qwen2-72B has downloadable open weights; the other is API-only.

## Snapshot

| | Claude Sonnet 4 | Qwen2-72B |
|---|---|---|
| Provider | Anthropic | Alibaba (Qwen) |
| Noometry Index | 40.6 | 29.5 |
| Rank | 139 | 294 |
| Context | 200K | — |
| Input $/M | $3 | — |
| Output $/M | $15 | — |
| Weights | Proprietary | Open |

## Coding

- Claude Sonnet 4: 44.0 (#70)
- Qwen2-72B: 28.7 (#303)

| Benchmark | Claude Sonnet 4 | Qwen2-72B |
|---|---|---|
| WeirdML | 46.1% | 11.3% |
| LMArena Coding | 1414 | 1196 |
| SWE-bench Verified (bash only) | 64.9% | — |
| Aider Polyglot | 61.3% | — |
| SciCode | 40% | — |
| GSO | 4.9% | — |
| BigCodeBench Instruct | — | 38.5% |
| BigCodeBench Complete | — | 54% |
| ALE-Bench | 655.35 | — |

## Agentic & Tool Use

- Claude Sonnet 4: 38.3 (#29)
- Qwen2-72B: 16.9 (#138)

| Benchmark | Claude Sonnet 4 | Qwen2-72B |
|---|---|---|
| TheAgentCompany | 33.1% | 1.1% |
| METR Time Horizons | 62% | 29.9% |
| Cybench | 35% | — |
| DeepResearch Bench | 46.6% | — |
| OSWorld | 43.9% | — |

## Reasoning

- Claude Sonnet 4: 23.0 (#169)
- Qwen2-72B: 22.2 (#184)

| Benchmark | Claude Sonnet 4 | Qwen2-72B |
|---|---|---|
| LMArena Hard Prompts | 1372 | 1191 |
| Epoch Capabilities Index | 141.69 | 125.27 |
| ARC-AGI-2 | 5.9% | — |
| SimpleBench | 45.5% | — |
| Kagi LLM Benchmark | 73% | — |
| ARC-AGI-1 | 40% | — |
| CritPt | 0.3% | — |
| EnigmaEval | 3.1% | — |
| DTBench | 77.1% | — |
| LMCA | 29% | — |
| ForecastBench | 60.2 | — |

## Math

- Claude Sonnet 4: 41.5 (#71)
- Qwen2-72B: 27.9 (#232)

| Benchmark | Claude Sonnet 4 | Qwen2-72B |
|---|---|---|
| LMArena Math | 1375 | 1235 |
| MATH Level 5 | 84.4% | 39.1% |
| OTIS Mock AIME 2024-2025 | 71.1% | — |
| Omni-MATH | 60.2% | — |
| FrontierMath (Feb 2025 set) | 4.1% | — |
| FrontierMath Tier 4 (v1) | 0% | — |

## Knowledge

- Claude Sonnet 4: 41.6 (#99)
- Qwen2-72B: 21.2 (#270)

| Benchmark | Claude Sonnet 4 | Qwen2-72B |
|---|---|---|
| GPQA Diamond | 79.2% | 40.8% |
| LMArena Expert | 1371 | 1171 |
| Humanity's Last Exam | 7.8% | — |
| MMLU-Pro | 84.3% | — |
| Confabulations | 13.2% | — |
| Vectara Hallucination Rate | 10.3% | — |
| GPQA (HELM) | 70.6% | — |
| MMLU | — | 82.4% |

## Multimodal

- Claude Sonnet 4: 25.6 (#116)
- Qwen2-72B: —

| Benchmark | Claude Sonnet 4 | Qwen2-72B |
|---|---|---|
| LMArena Vision | 1191 | — |
| GeoBench | 37% | — |
| VPCT | 34% | — |
| MindCube | 44.8% | — |

## Multilingual

- Claude Sonnet 4: 46.9 (#151)
- Qwen2-72B: 36.0 (#239)

| Benchmark | Claude Sonnet 4 | Qwen2-72B |
|---|---|---|
| LMArena Non-English | 1333 | 1176 |
| LMArena Chinese | 1351 | 1240 |
| LMArena French | 1364 | 1169 |
| LMArena German | 1331 | 1151 |
| LMArena Japanese | 1301 | 1111 |
| LMArena Korean | 1293 | 1083 |
| LMArena Russian | 1355 | 1169 |
| LMArena Spanish | 1356 | 1168 |

## Instruction Following

- Claude Sonnet 4: 71.8 (#141)
- Qwen2-72B: 62.0 (#236)

| Benchmark | Claude Sonnet 4 | Qwen2-72B |
|---|---|---|
| LMArena Instruction Following | 1375 | 1181 |
| IFEval | 84% | — |

## Long Context

- Claude Sonnet 4: 33.6 (#254)
- Qwen2-72B: 36.1 (#230)

| Benchmark | Claude Sonnet 4 | Qwen2-72B |
|---|---|---|
| LMArena Longer Query | 1398 | 1193 |
| Fiction.LiveBench | 46.9% | — |

## Writing & Preference

- Claude Sonnet 4: 57.3 (#127)
- Qwen2-72B: 41.0 (#236)

| Benchmark | Claude Sonnet 4 | Qwen2-72B |
|---|---|---|
| LMArena Text | 1351 | 1203 |
| LMArena Creative Writing | 1345 | 1181 |
| LMArena Multi-Turn | 1376 | 1195 |
| Short-Story Creative Writing | 81.4% | — |
| EQ-Bench Creative Writing | 1483 | — |
| WildBench | 83.8% | — |

## FAQ

### Is Claude Sonnet 4 better than Qwen2-72B?

Claude Sonnet 4 is the stronger model overall, scoring 40.6 to 29.5 on the Noometry Index.

### Is Claude Sonnet 4 or Qwen2-72B better for coding?

Claude Sonnet 4 scores higher on coding benchmarks: 44.0 versus 28.7 in the Noometry coding category.

### How many benchmarks do Claude Sonnet 4 and Qwen2-72B share?

23 benchmarks have published results for both models. Claude Sonnet 4 has 58 scored results on Noometry and Qwen2-72B has 26.