Model comparison
GPT-4o mini vs Qwen2.5-Coder-32B
Qwen2.5-Coder-32B is the stronger model overall, scoring 33.4 to 25.5 on the Noometry Index. GPT-4o mini costs 2.8× less per token, which makes it the better buy when Qwen2.5-Coder-32B's lead doesn't matter for your workload.
Last verified . 27 shared benchmarks.
Alibaba (Qwen)Summary
- They share 27 benchmarks with published results for both. GPT-4o mini scores higher in 3 categories and Qwen2.5-Coder-32B in 5 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen2.5-Coder-32B leads 33.3 to 10.4.
- The biggest single-benchmark swing is LiveBench Coding: 43.1% for GPT-4o mini and 56.9% for Qwen2.5-Coder-32B.
- GPT-4o mini is cheaper at $0.15 / $0.60 per million input/output tokens, against $0.66 / $1 for Qwen2.5-Coder-32B.
- GPT-4o mini accepts more context: 128K tokens versus 33K.
- Qwen2.5-Coder-32B has downloadable open weights; the other is API-only.
Side by side
| GPT-4o mini | Qwen2.5-Coder-32B | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 25.5 | 33.4 |
| Released | 2024-07-18 | 2024-09-18 |
| Weights | Proprietary | Open |
| Context window | 128K | 33K |
| Max output | 16K | 29K |
| Input $ / M tokens | $0.15 | $0.66 |
| Output $ / M tokens | $0.60 | $1 |
| Results tracked | 60 | 31 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
GPT-4o mini: 22.0 (#335), Qwen2.5-Coder-32B: 22.6 (#333)
| Benchmark | GPT-4o mini | Qwen2.5-Coder-32B |
|---|---|---|
| Aider Polyglot | 3.6% | 16.4% |
| BigCodeBench Instruct | 46.1% | 49% |
| LiveBench Coding | 43.1% | 56.9% |
| LMArena Coding | 1290 | 1276 |
| BigCodeBench Complete | 57.4% | 58% |
| HumanEval+ | 83.5% | 87.2% |
| MBPP+ | 72.2% | 77% |
| SWE-bench Verified (bash only) | — | 9% |
| WeirdML | 11.8% | — |
Agentic & Tool Use Not comparable
GPT-4o mini: 27.5 (#101), Qwen2.5-Coder-32B: —
| Benchmark | GPT-4o mini | Qwen2.5-Coder-32B |
|---|---|---|
| BALROG | 17.4% | — |
Reasoning Qwen2.5-Coder-32B leads
GPT-4o mini: 8.7 (#347), Qwen2.5-Coder-32B: 21.2 (#225)
| Benchmark | GPT-4o mini | Qwen2.5-Coder-32B |
|---|---|---|
| LiveBench Reasoning | 32.8% | 42.1% |
| LMArena Hard Prompts | 1267 | 1251 |
| LiveBench Data Analysis | 50% | 49.9% |
| Epoch Capabilities Index | 126.56 | 119.49 |
| LiveBench | 41.3% | 46.2% |
| ARC-AGI-2 | 0% | — |
| SimpleBench | 10.7% | — |
| Kagi LLM Benchmark | 28.8% | — |
| Chess Puzzles | 0% | — |
| Mystery Game Puzzles | 12% | — |
| DTBench | 54.4% | — |
| LMCA | 10.4% | — |
| HellaSwag | — | 83% |
| PIQA | 88.7% | — |
| WinoGrande | — | 80.8% |
Math Qwen2.5-Coder-32B leads
GPT-4o mini: 10.4 (#314), Qwen2.5-Coder-32B: 33.3 (#204)
| Benchmark | GPT-4o mini | Qwen2.5-Coder-32B |
|---|---|---|
| LiveBench Math | 36.3% | 46.6% |
| LMArena Math | 1267 | 1251 |
| GSM8K | 91.3% | 93% |
| FrontierMath (Tiers 1-3) | 0.7% | — |
| OTIS Mock AIME 2024-2025 | 6.9% | — |
| Omni-MATH | 28% | — |
| MATH Level 5 | 52.6% | — |
Knowledge Qwen2.5-Coder-32B leads
GPT-4o mini: 17.7 (#284), Qwen2.5-Coder-32B: 33.4 (#203)
| Benchmark | GPT-4o mini | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Expert | 1235 | 1221 |
| MMLU | 81.8% | 79.1% |
| GPQA Diamond | 37.7% | — |
| SimpleQA Verified | 8.3% | — |
| MMLU-Pro | 60.3% | — |
| Confabulations | 37.2% | — |
| GPQA (HELM) | 36.8% | — |
| ARC (AI2) Challenge | — | 70.5% |
| BoolQ | 88.7% | — |
Multimodal Not comparable
GPT-4o mini: 25.9 (#122), Qwen2.5-Coder-32B: —
| Benchmark | GPT-4o mini | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Vision | 1066 | — |
| Video-MME | 64.8% | — |
| GeoBench | 64% | — |
| VPCT | 34% | — |
Multilingual GPT-4o mini leads
GPT-4o mini: 42.0 (#199), Qwen2.5-Coder-32B: 37.8 (#235)
| Benchmark | GPT-4o mini | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Non-English | 1266 | 1205 |
| LMArena Chinese | 1265 | 1222 |
| LMArena Russian | 1275 | 1228 |
| LMArena French | 1297 | — |
| LMArena German | 1272 | — |
| LMArena Japanese | 1216 | — |
| LMArena Korean | 1195 | — |
| LMArena Spanish | 1276 | — |
Instruction Following Too close to call
GPT-4o mini: 61.9 (#239), Qwen2.5-Coder-32B: 61.4 (#245)
| Benchmark | GPT-4o mini | Qwen2.5-Coder-32B |
|---|---|---|
| LiveBench Instruction Following | 56.8% | 58.7% |
| LMArena Instruction Following | 1258 | 1223 |
| IFEval | 78.2% | — |
Long Context GPT-4o mini leads
GPT-4o mini: 39.1 (#186), Qwen2.5-Coder-32B: 38.0 (#208)
| Benchmark | GPT-4o mini | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Longer Query | 1289 | 1251 |
Writing & Preference Qwen2.5-Coder-32B leads
GPT-4o mini: 39.5 (#248), Qwen2.5-Coder-32B: 41.6 (#240)
| Benchmark | GPT-4o mini | Qwen2.5-Coder-32B |
|---|---|---|
| LMArena Text | 1286 | 1230 |
| LMArena Creative Writing | 1268 | 1174 |
| LMArena Multi-Turn | 1285 | 1222 |
| LiveBench Language | 28.6% | 23.3% |
| Short-Story Creative Writing | 67.2% | — |
| EQ-Bench Creative Writing | 873 | — |
| WildBench | 79.1% | — |
Frequently asked questions
Is GPT-4o mini better than Qwen2.5-Coder-32B?
Qwen2.5-Coder-32B is the stronger model overall, scoring 33.4 to 25.5 on the Noometry Index. GPT-4o mini costs 2.8× less per token, which makes it the better buy when Qwen2.5-Coder-32B's lead doesn't matter for your workload.
Which is cheaper, GPT-4o mini or Qwen2.5-Coder-32B?
GPT-4o mini is cheaper. It lists at $0.15 per million input tokens and $0.60 per million output tokens; Qwen2.5-Coder-32B lists at $0.66 and $1.
Is GPT-4o mini or Qwen2.5-Coder-32B better for coding?
They score almost the same on coding (22.0 vs 22.6); test both on your own repository before choosing.
Which has the bigger context window?
GPT-4o mini does, with 128K tokens against 33K.
How many benchmarks do GPT-4o mini and Qwen2.5-Coder-32B share?
27 benchmarks have published results for both models. GPT-4o mini has 60 scored results on Noometry and Qwen2.5-Coder-32B has 31.