Model comparison
Claude Sonnet 4 vs Qwen3.5 397B-A17B
Qwen3.5 397B-A17B is the stronger model overall, scoring 46.0 to 40.8 on the Noometry Index.
Last verified . 25 shared benchmarks.
Alibaba (Qwen)Summary
- They share 25 benchmarks with published results for both. Claude Sonnet 4 scores higher in 2 categories and Qwen3.5 397B-A17B in 8 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in multimodal, where Qwen3.5 397B-A17B leads 40.7 to 26.2.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 71.1% for Claude Sonnet 4 and 88.9% for Qwen3.5 397B-A17B.
- Qwen3.5 397B-A17B is cheaper at $0.60 / $3.60 per million input/output tokens, against $3 / $15 for Claude Sonnet 4.
- Qwen3.5 397B-A17B accepts more context: 262K tokens versus 200K.
- Qwen3.5 397B-A17B has downloadable open weights; the other is API-only.
Side by side
| Claude Sonnet 4 | Qwen3.5 397B-A17B | |
|---|---|---|
| Provider | Anthropic | Alibaba (Qwen) |
| Noometry Index | 40.8 | 46.0 |
| Released | 2025-05-22 | 2026-02-01 |
| Weights | Proprietary | Open |
| Context window | 200K | 262K |
| Max output | 64K | 66K |
| Input $ / M tokens | $3 | $0.60 |
| Output $ / M tokens | $15 | $3.60 |
| Results tracked | 58 | 36 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Sonnet 4 leads
Claude Sonnet 4: 43.5 (#88), Qwen3.5 397B-A17B: 42.0 (#114)
| Benchmark | Claude Sonnet 4 | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Coding | 1414 | 1465 |
| SWE-bench Verified (bash only) | 64.9% | — |
| Aider Polyglot | 61.3% | — |
| LMArena WebDev | — | 1400 |
| SciCode | 40% | — |
| GSO | 4.9% | — |
| WeirdML | 46.1% | — |
| ALE-Bench | 655.35 | — |
Agentic & Tool Use Claude Sonnet 4 leads
Claude Sonnet 4: 38.5 (#31), Qwen3.5 397B-A17B: 33.3 (#53)
| Benchmark | Claude Sonnet 4 | Qwen3.5 397B-A17B |
|---|---|---|
| APEX-Agents | — | 24.9% |
| TheAgentCompany | 33.1% | — |
| τ²-bench Airline | — | 81.5% |
| τ²-bench Banking | — | 9.8% |
| τ²-bench Retail | — | 84.4% |
| τ²-bench Telecom | — | 97.8% |
| Cybench | 35% | — |
| DeepResearch Bench | 46.6% | — |
| OSWorld | 43.9% | — |
| METR Time Horizons | 62% | — |
Reasoning Qwen3.5 397B-A17B leads
Claude Sonnet 4: 22.9 (#187), Qwen3.5 397B-A17B: 34.5 (#70)
| Benchmark | Claude Sonnet 4 | Qwen3.5 397B-A17B |
|---|---|---|
| Kagi LLM Benchmark | 73% | 73.7% |
| LMArena Hard Prompts | 1372 | 1448 |
| DTBench | 77.1% | 87.5% |
| LMCA | 29% | 37.9% |
| Epoch Capabilities Index | 141.69 | 146.65 |
| ARC-AGI-2 | 5.9% | — |
| SimpleBench | 45.5% | — |
| NYT Connections (extended) | — | 58.9% |
| ARC-AGI-1 | 40% | — |
| CritPt | 0.3% | — |
| Chess Puzzles | — | 13% |
| EnigmaEval | 3.1% | — |
| Thematic Generalization | — | 65.1% |
| Mystery Game Puzzles | — | 18% |
| ForecastBench | 60.2 | — |
Math Qwen3.5 397B-A17B leads
Claude Sonnet 4: 43.3 (#80), Qwen3.5 397B-A17B: 46.1 (#73)
| Benchmark | Claude Sonnet 4 | Qwen3.5 397B-A17B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 71.1% | 88.9% |
| LMArena Math | 1375 | 1454 |
| FrontierMath (Tiers 1-3) | — | 31.2% |
| Omni-MATH | 60.2% | — |
| MATH Level 5 | 84.4% | — |
| FrontierMath (Feb 2025 set) | 4.1% | — |
| FrontierMath Tier 4 (v1) | 0% | — |
Knowledge Qwen3.5 397B-A17B leads
Claude Sonnet 4: 41.8 (#108), Qwen3.5 397B-A17B: 53.3 (#58)
| Benchmark | Claude Sonnet 4 | Qwen3.5 397B-A17B |
|---|---|---|
| GPQA Diamond | 79.2% | 86.4% |
| LMArena Expert | 1372 | 1462 |
| Humanity's Last Exam | 7.8% | — |
| MMLU-Pro | 84.3% | — |
| Confabulations | 13.2% | — |
| Vectara Hallucination Rate | 10.3% | — |
| GPQA (HELM) | 70.6% | — |
Multimodal Qwen3.5 397B-A17B leads
Claude Sonnet 4: 26.2 (#121), Qwen3.5 397B-A17B: 40.7 (#44)
| Benchmark | Claude Sonnet 4 | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Vision | 1191 | 1263 |
| GeoBench | 37% | — |
| VPCT | 34% | — |
| MindCube | 44.8% | — |
Multilingual Qwen3.5 397B-A17B leads
Claude Sonnet 4: 46.7 (#156), Qwen3.5 397B-A17B: 53.7 (#59)
| Benchmark | Claude Sonnet 4 | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Non-English | 1333 | 1430 |
| LMArena Chinese | 1350 | 1500 |
| LMArena French | 1363 | 1461 |
| LMArena German | 1331 | 1447 |
| LMArena Japanese | 1302 | 1426 |
| LMArena Korean | 1291 | 1384 |
| LMArena Russian | 1355 | 1429 |
| LMArena Spanish | 1357 | 1441 |
Instruction Following Qwen3.5 397B-A17B leads
Claude Sonnet 4: 71.7 (#145), Qwen3.5 397B-A17B: 75.0 (#77)
| Benchmark | Claude Sonnet 4 | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Instruction Following | 1376 | 1424 |
| IFEval | 84% | — |
Long Context Qwen3.5 397B-A17B leads
Claude Sonnet 4: 33.7 (#259), Qwen3.5 397B-A17B: 44.1 (#74)
| Benchmark | Claude Sonnet 4 | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Longer Query | 1398 | 1442 |
| Fiction.LiveBench | 46.9% | — |
Writing & Preference Qwen3.5 397B-A17B leads
Claude Sonnet 4: 57.1 (#132), Qwen3.5 397B-A17B: 62.3 (#79)
| Benchmark | Claude Sonnet 4 | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Text | 1351 | 1438 |
| LMArena Creative Writing | 1345 | 1401 |
| EQ-Bench Creative Writing | 1483 | 1478 |
| LMArena Multi-Turn | 1376 | 1446 |
| Short-Story Creative Writing | 81.4% | — |
| WildBench | 83.8% | — |
Frequently asked questions
Is Claude Sonnet 4 better than Qwen3.5 397B-A17B?
Qwen3.5 397B-A17B is the stronger model overall, scoring 46.0 to 40.8 on the Noometry Index.
Which is cheaper, Claude Sonnet 4 or Qwen3.5 397B-A17B?
Qwen3.5 397B-A17B is cheaper. It lists at $0.60 per million input tokens and $3.60 per million output tokens; Claude Sonnet 4 lists at $3 and $15.
Is Claude Sonnet 4 or Qwen3.5 397B-A17B better for coding?
Claude Sonnet 4 scores higher on coding benchmarks: 43.5 versus 42.0 in the Noometry coding category.
Which has the bigger context window?
Qwen3.5 397B-A17B does, with 262K tokens against 200K.
How many benchmarks do Claude Sonnet 4 and Qwen3.5 397B-A17B share?
25 benchmarks have published results for both models. Claude Sonnet 4 has 58 scored results on Noometry and Qwen3.5 397B-A17B has 36.