Model comparison
GPT-4.1 vs Qwen2.5-Coder-32B
GPT-4.1 is the stronger model overall, scoring 35.9 to 33.4 on the Noometry Index. Qwen2.5-Coder-32B costs 4.7× less per token, which makes it the better buy when GPT-4.1's lead doesn't matter for your workload.
Last verified . 15 shared benchmarks.
Alibaba (Qwen)Summary
- They share 15 benchmarks with published results for both. GPT-4.1 scores higher in 6 categories and Qwen2.5-Coder-32B in 2 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where GPT-4.1 leads 57.6 to 41.6.
- The biggest single-benchmark swing is Aider Polyglot: 52.4% for GPT-4.1 and 16.4% for Qwen2.5-Coder-32B.
- Qwen2.5-Coder-32B is cheaper at $0.66 / $1 per million input/output tokens, against $2 / $8 for GPT-4.1.
- GPT-4.1 accepts more context: 1.05M tokens versus 33K.
- Qwen2.5-Coder-32B has downloadable open weights; the other is API-only.
Side by side
| GPT-4.1 | Qwen2.5-Coder-32B | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 35.9 | 33.4 |
| Released | 2025-04-14 | 2024-09-18 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 33K |
| Max output | 33K | 29K |
| Input $ / M tokens | $2 | $0.66 |
| Output $ / M tokens | $8 | $1 |
| Results tracked | 52 | 31 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-4.1 leads
GPT-4.1: 34.4 (#238), Qwen2.5-Coder-32B: 22.6 (#333)
| Benchmark | GPT-4.1 | Qwen2.5-Coder-32B |
|---|---|---|
| SWE-bench Verified (bash only) | 39.6% | 9% |
| Aider Polyglot | 52.4% | 16.4% |
| LMArena Coding | 1391 | 1276 |
| SWE-bench Verified | 48.5% | — |
| WeirdML | 39% | — |
| BigCodeBench Instruct | — | 49% |
| LiveBench Coding | — | 56.9% |
| BigCodeBench Complete | — | 58% |
| CadEval | 42% | — |
| ALE-Bench | 558.1 | — |
| HumanEval+ | — | 87.2% |
| MBPP+ | — | 77% |
Agentic & Tool Use Not comparable
GPT-4.1: 34.7 (#43), Qwen2.5-Coder-32B: —
| Benchmark | GPT-4.1 | Qwen2.5-Coder-32B |
|---|---|---|
| Berkeley Function Calling Leaderboard | 54% | — |
Reasoning Qwen2.5-Coder-32B leads
GPT-4.1: 11.7 (#339), Qwen2.5-Coder-32B: 21.2 (#225)
| Benchmark | GPT-4.1 | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Hard Prompts | 1384 | 1251 |
| Epoch Capabilities Index | 136.78 | 119.49 |
| ARC-AGI-2 | 0.4% | — |
| SimpleBench | 27% | — |
| Kagi LLM Benchmark | 52.3% | — |
| ARC-AGI-1 | 5.5% | — |
| Chess Puzzles | 6% | — |
| EnigmaEval | 2.2% | — |
| LiveBench Reasoning | — | 42.1% |
| DTBench | 68.3% | — |
| LiveBench Data Analysis | — | 49.9% |
| LMCA | 25.6% | — |
| ForecastBench | 61.5 | — |
| HellaSwag | — | 83% |
| LiveBench | — | 46.2% |
| WinoGrande | — | 80.8% |
Math Qwen2.5-Coder-32B leads
GPT-4.1: 22.3 (#280), Qwen2.5-Coder-32B: 33.3 (#204)
| Benchmark | GPT-4.1 | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Math | 1370 | 1251 |
| FrontierMath (Tiers 1-3) | 6% | — |
| OTIS Mock AIME 2024-2025 | 38.3% | — |
| Omni-MATH | 47.1% | — |
| LiveBench Math | — | 46.6% |
| MATH Level 5 | 83% | — |
| FrontierMath (Feb 2025 set) | 5.5% | — |
| FrontierMath Tier 4 (v1) | 0% | — |
| GSM8K | — | 93% |
Knowledge GPT-4.1 leads
GPT-4.1: 37.1 (#160), Qwen2.5-Coder-32B: 33.4 (#203)
| Benchmark | GPT-4.1 | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Expert | 1364 | 1221 |
| GPQA Diamond | 66.9% | — |
| Humanity's Last Exam | 5.4% | — |
| SimpleQA Verified | 31.1% | — |
| MMLU-Pro | 81.1% | — |
| Vectara Hallucination Rate | 5.6% | — |
| GPQA (HELM) | 65.9% | — |
| ARC (AI2) Challenge | — | 70.5% |
| MMLU | — | 79.1% |
Multimodal Not comparable
GPT-4.1: 38.2 (#67), Qwen2.5-Coder-32B: —
| Benchmark | GPT-4.1 | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Vision | 1211 | — |
| GeoBench | 72% | — |
Multilingual GPT-4.1 leads
GPT-4.1: 49.4 (#133), Qwen2.5-Coder-32B: 37.8 (#235)
| Benchmark | GPT-4.1 | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Non-English | 1370 | 1205 |
| LMArena Chinese | 1382 | 1222 |
| LMArena Russian | 1377 | 1228 |
| LMArena French | 1382 | — |
| LMArena German | 1381 | — |
| LMArena Japanese | 1319 | — |
| LMArena Korean | 1339 | — |
| LMArena Spanish | 1376 | — |
Instruction Following GPT-4.1 leads
GPT-4.1: 71.3 (#153), Qwen2.5-Coder-32B: 61.4 (#245)
| Benchmark | GPT-4.1 | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Instruction Following | 1367 | 1223 |
| LiveBench Instruction Following | — | 58.7% |
| IFEval | 83.8% | — |
Long Context GPT-4.1 leads
GPT-4.1: 40.0 (#163), Qwen2.5-Coder-32B: 38.0 (#208)
| Benchmark | GPT-4.1 | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Longer Query | 1385 | 1251 |
| Fiction.LiveBench | 63.9% | — |
Writing & Preference GPT-4.1 leads
GPT-4.1: 57.6 (#125), Qwen2.5-Coder-32B: 41.6 (#240)
| Benchmark | GPT-4.1 | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Text | 1383 | 1230 |
| LMArena Creative Writing | 1363 | 1174 |
| LMArena Multi-Turn | 1398 | 1222 |
| EQ-Bench Creative Writing | 1420 | — |
| WildBench | 85.4% | — |
| LiveBench Language | — | 23.3% |
Frequently asked questions
Is GPT-4.1 better than Qwen2.5-Coder-32B?
GPT-4.1 is the stronger model overall, scoring 35.9 to 33.4 on the Noometry Index. Qwen2.5-Coder-32B costs 4.7× less per token, which makes it the better buy when GPT-4.1's lead doesn't matter for your workload.
Which is cheaper, GPT-4.1 or Qwen2.5-Coder-32B?
Qwen2.5-Coder-32B is cheaper. It lists at $0.66 per million input tokens and $1 per million output tokens; GPT-4.1 lists at $2 and $8.
Is GPT-4.1 or Qwen2.5-Coder-32B better for coding?
GPT-4.1 scores higher on coding benchmarks: 34.4 versus 22.6 in the Noometry coding category.
Which has the bigger context window?
GPT-4.1 does, with 1.05M tokens against 33K.
How many benchmarks do GPT-4.1 and Qwen2.5-Coder-32B share?
15 benchmarks have published results for both models. GPT-4.1 has 52 scored results on Noometry and Qwen2.5-Coder-32B has 31.