Model comparison
Claude Sonnet 4.5 vs Qwen3.8 Max
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 44.1 on the Noometry Index.
Last verified . 33 shared benchmarks.
Alibaba (Qwen)Summary
- They share 33 benchmarks with published results for both. Claude Sonnet 4.5 scores higher in 0 categories and Qwen3.8 Max in 10 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3.8 Max leads 73.2 to 32.3.
- The biggest single-benchmark swing is NYT Connections (extended): 37.3% for Claude Sonnet 4.5 and 88.3% for Qwen3.8 Max.
- Qwen3.8 Max is cheaper at $2 / $6 per million input/output tokens, against $3 / $15 for Claude Sonnet 4.5.
- Qwen3.8 Max accepts more context: 1M tokens versus 200K.
Side by side
| Claude Sonnet 4.5 | Qwen3.8 Max | |
|---|---|---|
| Provider | Anthropic | Alibaba (Qwen) |
| Noometry Index | 44.1 | 56.8 |
| Released | 2025-09-29 | 2026-08-02 |
| Weights | Proprietary | Proprietary |
| Context window | 200K | 1M |
| Max output | 64K | 131K |
| Input $ / M tokens | $3 | $2 |
| Output $ / M tokens | $15 | $6 |
| Results tracked | 73 | 39 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.8 Max leads
Claude Sonnet 4.5: 47.3 (#61), Qwen3.8 Max: 53.5 (#29)
| Benchmark | Claude Sonnet 4.5 | Qwen3.8 Max |
|---|---|---|
| LMArena WebDev | 1393 | 1674 |
| SciCode | 44.7% | 53.2% |
| LMArena Coding | 1489 | 1502 |
| SWE-bench Verified | 71.3% | — |
| DeepSWE | — | 57.5% |
| SWE-bench Verified (bash only) | 71.4% | — |
| SWE-bench Multilingual | 67% | — |
| FrontierSWE | — | 17.8% |
| GSO | 14.7% | — |
| WeirdML | 47.7% | — |
| ALE-Bench | 796.15 | — |
| AlgoTune | 1.52 | — |
Agentic & Tool Use Qwen3.8 Max leads
Claude Sonnet 4.5: 38.3 (#32), Qwen3.8 Max: 45.4 (#14)
| Benchmark | Claude Sonnet 4.5 | Qwen3.8 Max |
|---|---|---|
| τ²-bench Banking | 25.3% | 55.1% |
| Terminal-Bench | 46.5% | — |
| APEX-Agents | — | 63.3% |
| Berkeley Function Calling Leaderboard | 73.2% | — |
| GDPval | 42.5% | — |
| Remote Labor Index | 2.1% | — |
| τ²-bench Airline | 72% | — |
| τ²-bench Retail | 72.4% | — |
| τ²-bench Telecom | 84.9% | — |
| Cybench | 60% | — |
| DeepResearch Bench | 52.6% | — |
| OSWorld | 62.9% | — |
| GDP.pdf | — | 23.2% |
| LMArena Search | 1159 | — |
| METR Time Horizons | 67.4% | — |
| Vending-Bench 2 | 3,839 | — |
Reasoning Qwen3.8 Max leads
Claude Sonnet 4.5: 26.9 (#125), Qwen3.8 Max: 54.4 (#26)
| Benchmark | Claude Sonnet 4.5 | Qwen3.8 Max |
|---|---|---|
| NYT Connections (extended) | 37.3% | 88.3% |
| CritPt | 1.1% | 20% |
| Chess Puzzles | 12% | 40% |
| LMArena Hard Prompts | 1462 | 1496 |
| Mystery Game Puzzles | 17% | 38% |
| DTBench | 83.2% | 92% |
| LMCA | 38.8% | 46.2% |
| Epoch Capabilities Index | 146.84 | 156.41 |
| ARC-AGI-2 | 13.6% | — |
| SimpleBench | 54.3% | — |
| Kagi LLM Benchmark | 57.9% | — |
| ARC-AGI-1 | 63.7% | — |
| EnigmaEval | 6% | — |
| EBR-Bench | 2.4% | — |
| ForecastBench | 61.9 | — |
Math Qwen3.8 Max leads
Claude Sonnet 4.5: 32.3 (#216), Qwen3.8 Max: 73.2 (#20)
| Benchmark | Claude Sonnet 4.5 | Qwen3.8 Max |
|---|---|---|
| FrontierMath (Tiers 1-3) | 23.9% | 74.7% |
| FrontierMath Tier 4 | 2.4% | 46.3% |
| OTIS Mock AIME 2024-2025 | 77.8% | 100% |
| ProofBench | 19% | 58% |
| LMArena Math | 1449 | 1499 |
| Omni-MATH | 55.3% | — |
| MATH Level 5 | 97.7% | — |
| FrontierMath (Feb 2025 set) | 15.2% | — |
| FrontierMath Tier 4 (v1) | 4.2% | — |
Knowledge Qwen3.8 Max leads
Claude Sonnet 4.5: 48.4 (#76), Qwen3.8 Max: 61.7 (#27)
| Benchmark | Claude Sonnet 4.5 | Qwen3.8 Max |
|---|---|---|
| GPQA Diamond | 82.3% | 92.7% |
| SimpleQA Verified | 30.7% | 47.3% |
| LMArena Expert | 1482 | 1507 |
| Humanity's Last Exam | 13.7% | — |
| MMLU-Pro | 86.9% | — |
| Vectara Hallucination Rate | 12% | — |
| GPQA (HELM) | 68.6% | — |
Multimodal Qwen3.8 Max leads
Claude Sonnet 4.5: 34.8 (#89), Qwen3.8 Max: 37.2 (#75)
| Benchmark | Claude Sonnet 4.5 | Qwen3.8 Max |
|---|---|---|
| LMArena Vision | — | 1314 |
| VPCT | 39.8% | — |
| Furniture Assembly | — | 20% |
| LMArena Document | 1450 | — |
Multilingual Qwen3.8 Max leads
Claude Sonnet 4.5: 53.4 (#69), Qwen3.8 Max: 56.7 (#18)
| Benchmark | Claude Sonnet 4.5 | Qwen3.8 Max |
|---|---|---|
| LMArena Non-English | 1425 | 1472 |
| LMArena Chinese | 1459 | 1538 |
| LMArena French | 1458 | 1503 |
| LMArena German | 1427 | 1483 |
| LMArena Japanese | 1390 | 1467 |
| LMArena Korean | 1403 | 1461 |
| LMArena Russian | 1437 | 1481 |
| LMArena Spanish | 1457 | 1492 |
Instruction Following Qwen3.8 Max leads
Claude Sonnet 4.5: 75.0 (#78), Qwen3.8 Max: 77.6 (#17)
| Benchmark | Claude Sonnet 4.5 | Qwen3.8 Max |
|---|---|---|
| LMArena Instruction Following | 1459 | 1479 |
| IFEval | 85% | — |
Long Context Too close to call
Claude Sonnet 4.5: 45.2 (#46), Qwen3.8 Max: 45.6 (#31)
| Benchmark | Claude Sonnet 4.5 | Qwen3.8 Max |
|---|---|---|
| LMArena Longer Query | 1476 | 1489 |
Writing & Preference Too close to call
Claude Sonnet 4.5: 66.5 (#34), Qwen3.8 Max: 67.1 (#30)
| Benchmark | Claude Sonnet 4.5 | Qwen3.8 Max |
|---|---|---|
| LMArena Text | 1439 | 1483 |
| LMArena Creative Writing | 1442 | 1479 |
| LMArena Multi-Turn | 1465 | 1489 |
| EQ-Bench Creative Writing | 1678 | — |
| WildBench | 85.4% | — |
Frequently asked questions
Is Claude Sonnet 4.5 better than Qwen3.8 Max?
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 44.1 on the Noometry Index.
Which is cheaper, Claude Sonnet 4.5 or Qwen3.8 Max?
Qwen3.8 Max is cheaper. It lists at $2 per million input tokens and $6 per million output tokens; Claude Sonnet 4.5 lists at $3 and $15.
Is Claude Sonnet 4.5 or Qwen3.8 Max better for coding?
Qwen3.8 Max scores higher on coding benchmarks: 53.5 versus 47.3 in the Noometry coding category.
Which has the bigger context window?
Qwen3.8 Max does, with 1M tokens against 200K.
How many benchmarks do Claude Sonnet 4.5 and Qwen3.8 Max share?
33 benchmarks have published results for both models. Claude Sonnet 4.5 has 73 scored results on Noometry and Qwen3.8 Max has 39.