Model comparison
GPT-4 vs Qwen2.5-Max
Qwen2.5-Max is the stronger model overall, scoring 40.1 to 28.9 on the Noometry Index.
Last verified . 18 shared benchmarks.
Alibaba (Qwen)Summary
- They share 18 benchmarks with published results for both. GPT-4 scores higher in 0 categories and Qwen2.5-Max in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen2.5-Max leads 34.3 to 10.1.
Side by side
| GPT-4 | Qwen2.5-Max | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 28.9 | 40.1 |
| Released | 2023-03-14 | 2025-01-25 |
| Weights | Proprietary | Proprietary |
| Context window | 8K | — |
| Max output | 8K | — |
| Input $ / M tokens | $30 | — |
| Output $ / M tokens | $60 | — |
| Results tracked | 38 | 27 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen2.5-Max leads
GPT-4: 31.2 (#277), Qwen2.5-Max: 41.2 (#118)
| Benchmark | GPT-4 | Qwen2.5-Max |
|---|---|---|
| LMArena Coding | 1254 | 1359 |
| WeirdML | 12.4% | — |
| BigCodeBench Instruct | 46% | — |
| LiveBench Coding | — | 64.4% |
| BigCodeBench Complete | 57.2% | — |
| HumanEval+ | 79.3% | — |
Agentic & Tool Use Not comparable
GPT-4: —, Qwen2.5-Max: —
| Benchmark | GPT-4 | Qwen2.5-Max |
|---|---|---|
| METR Time Horizons | 36.1% | — |
Reasoning Qwen2.5-Max leads
GPT-4: 17.3 (#285), Qwen2.5-Max: 24.6 (#144)
| Benchmark | GPT-4 | Qwen2.5-Max |
|---|---|---|
| LMArena Hard Prompts | 1241 | 1360 |
| Epoch Capabilities Index | 125.88 | 132.54 |
| Chess Puzzles | 4% | — |
| LiveBench Reasoning | — | 51.4% |
| Mystery Game Puzzles | 12% | — |
| DTBench | 62.7% | — |
| LiveBench Data Analysis | — | 67.9% |
| LMCA | 17.1% | — |
| BIG-Bench Hard | 75.1% | — |
| ForecastBench | 57.8 | — |
| HellaSwag | 95.3% | — |
| LiveBench | — | 62.3% |
| WinoGrande | 87.5% | — |
Math Qwen2.5-Max leads
GPT-4: 10.1 (#304), Qwen2.5-Max: 34.3 (#162)
| Benchmark | GPT-4 | Qwen2.5-Max |
|---|---|---|
| LMArena Math | 1269 | 1369 |
| OTIS Mock AIME 2024-2025 | 1.1% | — |
| LiveBench Math | — | 58.4% |
| MATH Level 5 | 23% | — |
| GSM8K | 92% | — |
Knowledge Qwen2.5-Max leads
GPT-4: 18.4 (#277), Qwen2.5-Max: 34.8 (#181)
| Benchmark | GPT-4 | Qwen2.5-Max |
|---|---|---|
| LMArena Expert | 1211 | 1337 |
| GPQA Diamond | 35.7% | — |
| Confabulations | — | 21.8% |
| MMLU | 86.4% | — |
| TriviaQA | 84.8% | — |
Multilingual Qwen2.5-Max leads
GPT-4: 40.8 (#210), Qwen2.5-Max: 48.3 (#140)
| Benchmark | GPT-4 | Qwen2.5-Max |
|---|---|---|
| LMArena Non-English | 1246 | 1352 |
| LMArena Chinese | 1242 | 1382 |
| LMArena French | 1282 | 1396 |
| LMArena German | 1251 | 1350 |
| LMArena Japanese | 1209 | 1299 |
| LMArena Korean | 1184 | 1304 |
| LMArena Russian | 1251 | 1353 |
| LMArena Spanish | 1261 | 1377 |
Instruction Following Qwen2.5-Max leads
GPT-4: 65.5 (#217), Qwen2.5-Max: 71.5 (#147)
| Benchmark | GPT-4 | Qwen2.5-Max |
|---|---|---|
| LMArena Instruction Following | 1241 | 1335 |
| LiveBench Instruction Following | — | 75.3% |
Long Context Qwen2.5-Max leads
GPT-4: 37.7 (#207), Qwen2.5-Max: 41.3 (#137)
| Benchmark | GPT-4 | Qwen2.5-Max |
|---|---|---|
| LMArena Longer Query | 1244 | 1358 |
Writing & Preference Qwen2.5-Max leads
GPT-4: 35.0 (#263), Qwen2.5-Max: 55.6 (#141)
| Benchmark | GPT-4 | Qwen2.5-Max |
|---|---|---|
| LMArena Text | 1263 | 1367 |
| LMArena Creative Writing | 1244 | 1339 |
| LMArena Multi-Turn | 1257 | 1364 |
| Short-Story Creative Writing | — | 72.9% |
| EQ-Bench Creative Writing | 752 | — |
| LiveBench Language | — | 56.3% |
Frequently asked questions
Is GPT-4 better than Qwen2.5-Max?
Qwen2.5-Max is the stronger model overall, scoring 40.1 to 28.9 on the Noometry Index.
Is GPT-4 or Qwen2.5-Max better for coding?
Qwen2.5-Max scores higher on coding benchmarks: 41.2 versus 31.2 in the Noometry coding category.
How many benchmarks do GPT-4 and Qwen2.5-Max share?
18 benchmarks have published results for both models. GPT-4 has 38 scored results on Noometry and Qwen2.5-Max has 27.