Model comparison
GPT-4.1 vs Qwen3 235B-A22B
Qwen3 235B-A22B is the stronger model overall, scoring 43.5 to 35.9 on the Noometry Index.
Last verified . 43 shared benchmarks.
Alibaba (Qwen)Summary
- They share 43 benchmarks with published results for both. GPT-4.1 scores higher in 1 category and Qwen3 235B-A22B in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3 235B-A22B leads 50.4 to 22.3.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 38.3% for GPT-4.1 and 86.7% for Qwen3 235B-A22B.
- Qwen3 235B-A22B is cheaper at $0.70 / $2.80 per million input/output tokens, against $2 / $8 for GPT-4.1.
- GPT-4.1 accepts more context: 1.05M tokens versus 131K.
- Qwen3 235B-A22B has downloadable open weights; the other is API-only.
Side by side
| GPT-4.1 | Qwen3 235B-A22B | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 35.9 | 43.5 |
| Released | 2025-04-14 | 2025-04 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 131K |
| Max output | 33K | 16K |
| Input $ / M tokens | $2 | $0.70 |
| Output $ / M tokens | $8 | $2.80 |
| Results tracked | 52 | 49 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3 235B-A22B leads
GPT-4.1: 34.4 (#238), Qwen3 235B-A22B: 44.3 (#75)
| Benchmark | GPT-4.1 | Qwen3 235B-A22B |
|---|---|---|
| Aider Polyglot | 52.4% | 59.6% |
| WeirdML | 39% | 41% |
| LMArena Coding | 1391 | 1445 |
| SWE-bench Verified | 48.5% | — |
| SWE-bench Verified (bash only) | 39.6% | — |
| SciCode | — | 42.4% |
| CadEval | 42% | — |
| ALE-Bench | 558.1 | — |
Agentic & Tool Use Too close to call
GPT-4.1: 34.7 (#43), Qwen3 235B-A22B: 33.9 (#51)
| Benchmark | GPT-4.1 | Qwen3 235B-A22B |
|---|---|---|
| Berkeley Function Calling Leaderboard | 54% | 52.1% |
| Vending-Bench 2 | — | -11.34 |
Reasoning Qwen3 235B-A22B leads
GPT-4.1: 11.7 (#339), Qwen3 235B-A22B: 15.7 (#311)
| Benchmark | GPT-4.1 | Qwen3 235B-A22B |
|---|---|---|
| ARC-AGI-2 | 0.4% | 1.3% |
| SimpleBench | 27% | 31% |
| Kagi LLM Benchmark | 52.3% | 69.4% |
| ARC-AGI-1 | 5.5% | 11% |
| Chess Puzzles | 6% | 12% |
| LMArena Hard Prompts | 1384 | 1433 |
| DTBench | 68.3% | 80.3% |
| LMCA | 25.6% | 29.3% |
| Epoch Capabilities Index | 136.78 | 143.85 |
| ForecastBench | 61.5 | 59.7 |
| CritPt | — | 0% |
| EnigmaEval | 2.2% | — |
| Mystery Game Puzzles | — | 9% |
Math Qwen3 235B-A22B leads
GPT-4.1: 22.3 (#280), Qwen3 235B-A22B: 50.4 (#57)
| Benchmark | GPT-4.1 | Qwen3 235B-A22B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 38.3% | 86.7% |
| Omni-MATH | 47.1% | 71.8% |
| LMArena Math | 1370 | 1432 |
| MATH Level 5 | 83% | 68.9% |
| FrontierMath (Feb 2025 set) | 5.5% | 8.5% |
| FrontierMath Tier 4 (v1) | 0% | 0% |
| FrontierMath (Tiers 1-3) | 6% | — |
Knowledge Qwen3 235B-A22B leads
GPT-4.1: 37.1 (#160), Qwen3 235B-A22B: 49.6 (#73)
| Benchmark | GPT-4.1 | Qwen3 235B-A22B |
|---|---|---|
| GPQA Diamond | 66.9% | 80.1% |
| SimpleQA Verified | 31.1% | 40.4% |
| MMLU-Pro | 81.1% | 84.4% |
| Vectara Hallucination Rate | 5.6% | 9.3% |
| GPQA (HELM) | 65.9% | 72.7% |
| LMArena Expert | 1364 | 1463 |
| Humanity's Last Exam | 5.4% | — |
| Confabulations | — | 15.6% |
Multimodal Not comparable
GPT-4.1: 38.2 (#67), Qwen3 235B-A22B: —
| Benchmark | GPT-4.1 | Qwen3 235B-A22B |
|---|---|---|
| LMArena Vision | 1211 | — |
| GeoBench | 72% | — |
Multilingual Qwen3 235B-A22B leads
GPT-4.1: 49.4 (#133), Qwen3 235B-A22B: 52.3 (#89)
| Benchmark | GPT-4.1 | Qwen3 235B-A22B |
|---|---|---|
| LMArena Non-English | 1370 | 1409 |
| LMArena Chinese | 1382 | 1481 |
| LMArena French | 1382 | 1445 |
| LMArena German | 1381 | 1433 |
| LMArena Japanese | 1319 | 1399 |
| LMArena Korean | 1339 | 1391 |
| LMArena Russian | 1377 | 1411 |
| LMArena Spanish | 1376 | 1430 |
Instruction Following Qwen3 235B-A22B leads
GPT-4.1: 71.3 (#153), Qwen3 235B-A22B: 72.6 (#136)
| Benchmark | GPT-4.1 | Qwen3 235B-A22B |
|---|---|---|
| IFEval | 83.8% | 83.5% |
| LMArena Instruction Following | 1367 | 1408 |
Long Context Qwen3 235B-A22B leads
GPT-4.1: 40.0 (#163), Qwen3 235B-A22B: 46.1 (#26)
| Benchmark | GPT-4.1 | Qwen3 235B-A22B |
|---|---|---|
| Fiction.LiveBench | 63.9% | 75% |
| LMArena Longer Query | 1385 | 1426 |
Writing & Preference Qwen3 235B-A22B leads
GPT-4.1: 57.6 (#125), Qwen3 235B-A22B: 59.6 (#108)
| Benchmark | GPT-4.1 | Qwen3 235B-A22B |
|---|---|---|
| LMArena Text | 1383 | 1419 |
| LMArena Creative Writing | 1363 | 1384 |
| EQ-Bench Creative Writing | 1420 | 1366 |
| WildBench | 85.4% | 86.6% |
| LMArena Multi-Turn | 1398 | 1432 |
| Short-Story Creative Writing | — | 83% |
Frequently asked questions
Is GPT-4.1 better than Qwen3 235B-A22B?
Qwen3 235B-A22B is the stronger model overall, scoring 43.5 to 35.9 on the Noometry Index.
Which is cheaper, GPT-4.1 or Qwen3 235B-A22B?
Qwen3 235B-A22B is cheaper. It lists at $0.70 per million input tokens and $2.80 per million output tokens; GPT-4.1 lists at $2 and $8.
Is GPT-4.1 or Qwen3 235B-A22B better for coding?
Qwen3 235B-A22B scores higher on coding benchmarks: 44.3 versus 34.4 in the Noometry coding category.
Which has the bigger context window?
GPT-4.1 does, with 1.05M tokens against 131K.
How many benchmarks do GPT-4.1 and Qwen3 235B-A22B share?
43 benchmarks have published results for both models. GPT-4.1 has 52 scored results on Noometry and Qwen3 235B-A22B has 49.