Model comparison
GPT-4.1 vs Qwen3.8 Max
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 35.9 on the Noometry Index.
Last verified . 26 shared benchmarks.
Alibaba (Qwen)Summary
- They share 26 benchmarks with published results for both. GPT-4.1 scores higher in 1 category and Qwen3.8 Max in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3.8 Max leads 73.2 to 22.3.
- The biggest single-benchmark swing is FrontierMath (Tiers 1-3): 6% for GPT-4.1 and 74.7% for Qwen3.8 Max.
- Qwen3.8 Max is cheaper at $2 / $6 per million input/output tokens, against $2 / $8 for GPT-4.1.
- GPT-4.1 accepts more context: 1.05M tokens versus 1M.
Side by side
| GPT-4.1 | Qwen3.8 Max | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 35.9 | 56.8 |
| Released | 2025-04-14 | 2026-08-02 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 1M |
| Max output | 33K | 131K |
| Input $ / M tokens | $2 | $2 |
| Output $ / M tokens | $8 | $6 |
| Results tracked | 52 | 39 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.8 Max leads
GPT-4.1: 34.4 (#238), Qwen3.8 Max: 53.5 (#29)
| Benchmark | GPT-4.1 | Qwen3.8 Max |
|---|---|---|
| LMArena Coding | 1391 | 1502 |
| SWE-bench Verified | 48.5% | — |
| DeepSWE | — | 57.5% |
| SWE-bench Verified (bash only) | 39.6% | — |
| Aider Polyglot | 52.4% | — |
| LMArena WebDev | — | 1674 |
| FrontierSWE | — | 17.8% |
| SciCode | — | 53.2% |
| WeirdML | 39% | — |
| CadEval | 42% | — |
| ALE-Bench | 558.1 | — |
Agentic & Tool Use Qwen3.8 Max leads
GPT-4.1: 34.7 (#43), Qwen3.8 Max: 45.4 (#14)
| Benchmark | GPT-4.1 | Qwen3.8 Max |
|---|---|---|
| APEX-Agents | — | 63.3% |
| Berkeley Function Calling Leaderboard | 54% | — |
| τ²-bench Banking | — | 55.1% |
| GDP.pdf | — | 23.2% |
Reasoning Qwen3.8 Max leads
GPT-4.1: 11.7 (#339), Qwen3.8 Max: 54.4 (#26)
| Benchmark | GPT-4.1 | Qwen3.8 Max |
|---|---|---|
| Chess Puzzles | 6% | 40% |
| LMArena Hard Prompts | 1384 | 1496 |
| DTBench | 68.3% | 92% |
| LMCA | 25.6% | 46.2% |
| Epoch Capabilities Index | 136.78 | 156.41 |
| ARC-AGI-2 | 0.4% | — |
| SimpleBench | 27% | — |
| Kagi LLM Benchmark | 52.3% | — |
| NYT Connections (extended) | — | 88.3% |
| ARC-AGI-1 | 5.5% | — |
| CritPt | — | 20% |
| EnigmaEval | 2.2% | — |
| Mystery Game Puzzles | — | 38% |
| ForecastBench | 61.5 | — |
Math Qwen3.8 Max leads
GPT-4.1: 22.3 (#280), Qwen3.8 Max: 73.2 (#20)
| Benchmark | GPT-4.1 | Qwen3.8 Max |
|---|---|---|
| FrontierMath (Tiers 1-3) | 6% | 74.7% |
| OTIS Mock AIME 2024-2025 | 38.3% | 100% |
| LMArena Math | 1370 | 1499 |
| FrontierMath Tier 4 | — | 46.3% |
| ProofBench | — | 58% |
| Omni-MATH | 47.1% | — |
| MATH Level 5 | 83% | — |
| FrontierMath (Feb 2025 set) | 5.5% | — |
| FrontierMath Tier 4 (v1) | 0% | — |
Knowledge Qwen3.8 Max leads
GPT-4.1: 37.1 (#160), Qwen3.8 Max: 61.7 (#27)
| Benchmark | GPT-4.1 | Qwen3.8 Max |
|---|---|---|
| GPQA Diamond | 66.9% | 92.7% |
| SimpleQA Verified | 31.1% | 47.3% |
| LMArena Expert | 1364 | 1507 |
| Humanity's Last Exam | 5.4% | — |
| MMLU-Pro | 81.1% | — |
| Vectara Hallucination Rate | 5.6% | — |
| GPQA (HELM) | 65.9% | — |
Multimodal Too close to call
GPT-4.1: 38.2 (#67), Qwen3.8 Max: 37.2 (#75)
| Benchmark | GPT-4.1 | Qwen3.8 Max |
|---|---|---|
| LMArena Vision | 1211 | 1314 |
| GeoBench | 72% | — |
| Furniture Assembly | — | 20% |
Multilingual Qwen3.8 Max leads
GPT-4.1: 49.4 (#133), Qwen3.8 Max: 56.7 (#18)
| Benchmark | GPT-4.1 | Qwen3.8 Max |
|---|---|---|
| LMArena Non-English | 1370 | 1472 |
| LMArena Chinese | 1382 | 1538 |
| LMArena French | 1382 | 1503 |
| LMArena German | 1381 | 1483 |
| LMArena Japanese | 1319 | 1467 |
| LMArena Korean | 1339 | 1461 |
| LMArena Russian | 1377 | 1481 |
| LMArena Spanish | 1376 | 1492 |
Instruction Following Qwen3.8 Max leads
GPT-4.1: 71.3 (#153), Qwen3.8 Max: 77.6 (#17)
| Benchmark | GPT-4.1 | Qwen3.8 Max |
|---|---|---|
| LMArena Instruction Following | 1367 | 1479 |
| IFEval | 83.8% | — |
Long Context Qwen3.8 Max leads
GPT-4.1: 40.0 (#163), Qwen3.8 Max: 45.6 (#31)
| Benchmark | GPT-4.1 | Qwen3.8 Max |
|---|---|---|
| LMArena Longer Query | 1385 | 1489 |
| Fiction.LiveBench | 63.9% | — |
Writing & Preference Qwen3.8 Max leads
GPT-4.1: 57.6 (#125), Qwen3.8 Max: 67.1 (#30)
| Benchmark | GPT-4.1 | Qwen3.8 Max |
|---|---|---|
| LMArena Text | 1383 | 1483 |
| LMArena Creative Writing | 1363 | 1479 |
| LMArena Multi-Turn | 1398 | 1489 |
| EQ-Bench Creative Writing | 1420 | — |
| WildBench | 85.4% | — |
Frequently asked questions
Is GPT-4.1 better than Qwen3.8 Max?
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 35.9 on the Noometry Index.
Which is cheaper, GPT-4.1 or Qwen3.8 Max?
Qwen3.8 Max is cheaper. It lists at $2 per million input tokens and $6 per million output tokens; GPT-4.1 lists at $2 and $8.
Is GPT-4.1 or Qwen3.8 Max better for coding?
Qwen3.8 Max scores higher on coding benchmarks: 53.5 versus 34.4 in the Noometry coding category.
Which has the bigger context window?
GPT-4.1 does, with 1.05M tokens against 1M.
How many benchmarks do GPT-4.1 and Qwen3.8 Max share?
26 benchmarks have published results for both models. GPT-4.1 has 52 scored results on Noometry and Qwen3.8 Max has 39.