Model comparison
Claude Sonnet 4 vs Qwen2.5-VL 72B Instruct
Claude Sonnet 4 is the stronger model overall, scoring 40.8 to 29.9 on the Noometry Index.
Last verified . 4 shared benchmarks.
Alibaba (Qwen)Summary
- They share 4 benchmarks with published results for both. Claude Sonnet 4 scores higher in 2 categories and Qwen2.5-VL 72B Instruct in 1 category; 3 gaps are clear of the uncertainty.
- The widest gap is in agentic & tool use, where Claude Sonnet 4 leads 38.5 to 18.6.
- The biggest single-benchmark swing is OSWorld: 43.9% for Claude Sonnet 4 and 5% for Qwen2.5-VL 72B Instruct.
- Qwen2.5-VL 72B Instruct is cheaper at $2.80 / $8.40 per million input/output tokens, against $3 / $15 for Claude Sonnet 4.
- Claude Sonnet 4 accepts more context: 200K tokens versus 131K.
- Qwen2.5-VL 72B Instruct has downloadable open weights; the other is API-only.
Side by side
| Claude Sonnet 4 | Qwen2.5-VL 72B Instruct | |
|---|---|---|
| Provider | Anthropic | Alibaba (Qwen) |
| Noometry Index | 40.8 | 29.9 |
| Released | 2025-05-22 | 2024-09 |
| Weights | Proprietary | Open |
| Context window | 200K | 131K |
| Max output | 64K | 8K |
| Input $ / M tokens | $3 | $2.80 |
| Output $ / M tokens | $15 | $8.40 |
| Results tracked | 58 | 6 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Not comparable
Claude Sonnet 4: 43.5 (#88), Qwen2.5-VL 72B Instruct: —
| Benchmark | Claude Sonnet 4 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| SWE-bench Verified (bash only) | 64.9% | — |
| Aider Polyglot | 61.3% | — |
| SciCode | 40% | — |
| GSO | 4.9% | — |
| WeirdML | 46.1% | — |
| LMArena Coding | 1414 | — |
| ALE-Bench | 655.35 | — |
Agentic & Tool Use Claude Sonnet 4 leads
Claude Sonnet 4: 38.5 (#31), Qwen2.5-VL 72B Instruct: 18.6 (#144)
| Benchmark | Claude Sonnet 4 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| OSWorld | 43.9% | 5% |
| TheAgentCompany | 33.1% | — |
| Cybench | 35% | — |
| DeepResearch Bench | 46.6% | — |
| METR Time Horizons | 62% | — |
Reasoning Claude Sonnet 4 leads
Claude Sonnet 4: 22.9 (#187), Qwen2.5-VL 72B Instruct: 20.7 (#233)
| Benchmark | Claude Sonnet 4 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| Kagi LLM Benchmark | 73% | 36% |
| ARC-AGI-2 | 5.9% | — |
| SimpleBench | 45.5% | — |
| ARC-AGI-1 | 40% | — |
| CritPt | 0.3% | — |
| EnigmaEval | 3.1% | — |
| LMArena Hard Prompts | 1372 | — |
| DTBench | 77.1% | — |
| LMCA | 29% | — |
| Epoch Capabilities Index | 141.69 | — |
| ForecastBench | 60.2 | — |
Math Not comparable
Claude Sonnet 4: 43.3 (#80), Qwen2.5-VL 72B Instruct: —
| Benchmark | Claude Sonnet 4 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 71.1% | — |
| Omni-MATH | 60.2% | — |
| LMArena Math | 1375 | — |
| MATH Level 5 | 84.4% | — |
| FrontierMath (Feb 2025 set) | 4.1% | — |
| FrontierMath Tier 4 (v1) | 0% | — |
Knowledge Not comparable
Claude Sonnet 4: 41.8 (#108), Qwen2.5-VL 72B Instruct: —
| Benchmark | Claude Sonnet 4 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| GPQA Diamond | 79.2% | — |
| Humanity's Last Exam | 7.8% | — |
| MMLU-Pro | 84.3% | — |
| Confabulations | 13.2% | — |
| Vectara Hallucination Rate | 10.3% | — |
| GPQA (HELM) | 70.6% | — |
| LMArena Expert | 1372 | — |
Multimodal Qwen2.5-VL 72B Instruct leads
Claude Sonnet 4: 26.2 (#121), Qwen2.5-VL 72B Instruct: 33.5 (#97)
| Benchmark | Claude Sonnet 4 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| LMArena Vision | 1191 | 1107 |
| GeoBench | 37% | 62% |
| Video-MME | — | 73.5% |
| VPCT | 34% | — |
| MindCube | 44.8% | — |
| SpatialViz-Bench | — | 33.3% |
Multilingual Not comparable
Claude Sonnet 4: 46.7 (#156), Qwen2.5-VL 72B Instruct: —
| Benchmark | Claude Sonnet 4 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| LMArena Non-English | 1333 | — |
| LMArena Chinese | 1350 | — |
| LMArena French | 1363 | — |
| LMArena German | 1331 | — |
| LMArena Japanese | 1302 | — |
| LMArena Korean | 1291 | — |
| LMArena Russian | 1355 | — |
| LMArena Spanish | 1357 | — |
Instruction Following Not comparable
Claude Sonnet 4: 71.7 (#145), Qwen2.5-VL 72B Instruct: —
| Benchmark | Claude Sonnet 4 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| IFEval | 84% | — |
| LMArena Instruction Following | 1376 | — |
Long Context Not comparable
Claude Sonnet 4: 33.7 (#259), Qwen2.5-VL 72B Instruct: —
| Benchmark | Claude Sonnet 4 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| Fiction.LiveBench | 46.9% | — |
| LMArena Longer Query | 1398 | — |
Writing & Preference Not comparable
Claude Sonnet 4: 57.1 (#132), Qwen2.5-VL 72B Instruct: —
| Benchmark | Claude Sonnet 4 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| LMArena Text | 1351 | — |
| LMArena Creative Writing | 1345 | — |
| Short-Story Creative Writing | 81.4% | — |
| EQ-Bench Creative Writing | 1483 | — |
| WildBench | 83.8% | — |
| LMArena Multi-Turn | 1376 | — |
Frequently asked questions
Is Claude Sonnet 4 better than Qwen2.5-VL 72B Instruct?
Claude Sonnet 4 is the stronger model overall, scoring 40.8 to 29.9 on the Noometry Index.
Which is cheaper, Claude Sonnet 4 or Qwen2.5-VL 72B Instruct?
Qwen2.5-VL 72B Instruct is cheaper. It lists at $2.80 per million input tokens and $8.40 per million output tokens; Claude Sonnet 4 lists at $3 and $15.
Which has the bigger context window?
Claude Sonnet 4 does, with 200K tokens against 131K.
How many benchmarks do Claude Sonnet 4 and Qwen2.5-VL 72B Instruct share?
4 benchmarks have published results for both models. Claude Sonnet 4 has 58 scored results on Noometry and Qwen2.5-VL 72B Instruct has 6.