Model comparison
GPT-4.5 vs Qwen2.5-Max
Qwen2.5-Max is the stronger model overall, scoring 40.7 to 37.2 on the Noometry Index.
Last verified . 26 shared benchmarks.
Alibaba (Qwen)Summary
- They share 26 benchmarks with published results for both. GPT-4.5 scores higher in 4 categories and Qwen2.5-Max in 4 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Qwen2.5-Max leads 25.6 to 13.9.
- The biggest single-benchmark swing is LiveBench Reasoning: 71.1% for GPT-4.5 and 51.4% for Qwen2.5-Max.
Side by side
| GPT-4.5 | Qwen2.5-Max | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 37.2 | 40.7 |
| Released | 2025-02-27 | 2025-01-25 |
| Weights | Proprietary | Proprietary |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 42 | 27 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
GPT-4.5: 42.2 (#109), Qwen2.5-Max: 41.8 (#117)
| Benchmark | GPT-4.5 | Qwen2.5-Max |
|---|---|---|
| LiveBench Coding | 75.2% | 64.4% |
| LMArena Coding | 1396 | 1359 |
| Aider Polyglot | 44.9% | — |
| WeirdML | 39.4% | — |
Agentic & Tool Use Not comparable
GPT-4.5: 27.9 (#97), Qwen2.5-Max: —
| Benchmark | GPT-4.5 | Qwen2.5-Max |
|---|---|---|
| Cybench | 17.5% | — |
Reasoning Qwen2.5-Max leads
GPT-4.5: 13.9 (#330), Qwen2.5-Max: 25.6 (#147)
| Benchmark | GPT-4.5 | Qwen2.5-Max |
|---|---|---|
| LiveBench Reasoning | 71.1% | 51.4% |
| LMArena Hard Prompts | 1403 | 1360 |
| LiveBench Data Analysis | 64.3% | 67.9% |
| Epoch Capabilities Index | 136.74 | 132.53 |
| LiveBench | 69% | 62.3% |
| ARC-AGI-2 | 0.8% | — |
| SimpleBench | 34.5% | — |
| ARC-AGI-1 | 10.3% | — |
| EnigmaEval | 3.2% | — |
| ForecastBench | 61.7 | — |
Math Qwen2.5-Max leads
GPT-4.5: 32.6 (#211), Qwen2.5-Max: 36.9 (#162)
| Benchmark | GPT-4.5 | Qwen2.5-Max |
|---|---|---|
| LiveBench Math | 69.3% | 58.4% |
| LMArena Math | 1412 | 1369 |
| OTIS Mock AIME 2024-2025 | 37.8% | — |
| MATH Level 5 | 78.6% | — |
Knowledge Qwen2.5-Max leads
GPT-4.5: 32.5 (#211), Qwen2.5-Max: 35.3 (#186)
| Benchmark | GPT-4.5 | Qwen2.5-Max |
|---|---|---|
| Confabulations | 13.6% | 21.8% |
| LMArena Expert | 1394 | 1337 |
| GPQA Diamond | 68.7% | — |
| Humanity's Last Exam | 5.4% | — |
Multimodal Not comparable
GPT-4.5: 37.6 (#71), Qwen2.5-Max: —
| Benchmark | GPT-4.5 | Qwen2.5-Max |
|---|---|---|
| LMArena Vision | 1195 | — |
| VPCT | 45% | — |
Multilingual GPT-4.5 leads
GPT-4.5: 52.5 (#83), Qwen2.5-Max: 48.1 (#146)
| Benchmark | GPT-4.5 | Qwen2.5-Max |
|---|---|---|
| LMArena Non-English | 1413 | 1352 |
| LMArena Chinese | 1421 | 1382 |
| LMArena French | 1418 | 1396 |
| LMArena German | 1457 | 1350 |
| LMArena Japanese | 1416 | 1300 |
| LMArena Korean | 1392 | 1304 |
| LMArena Russian | 1419 | 1353 |
| LMArena Spanish | — | 1377 |
Instruction Following GPT-4.5 leads
GPT-4.5: 72.6 (#134), Qwen2.5-Max: 71.3 (#152)
| Benchmark | GPT-4.5 | Qwen2.5-Max |
|---|---|---|
| LiveBench Instruction Following | 72.3% | 75.3% |
| LMArena Instruction Following | 1404 | 1335 |
Long Context Too close to call
GPT-4.5: 40.4 (#155), Qwen2.5-Max: 41.4 (#142)
| Benchmark | GPT-4.5 | Qwen2.5-Max |
|---|---|---|
| LMArena Longer Query | 1406 | 1358 |
| Fiction.LiveBench | 63.9% | — |
Writing & Preference GPT-4.5 leads
GPT-4.5: 56.9 (#134), Qwen2.5-Max: 55.4 (#146)
| Benchmark | GPT-4.5 | Qwen2.5-Max |
|---|---|---|
| LMArena Text | 1417 | 1367 |
| LMArena Creative Writing | 1394 | 1339 |
| Short-Story Creative Writing | 75.6% | 72.9% |
| LMArena Multi-Turn | 1444 | 1364 |
| LiveBench Language | 61.5% | 56.3% |
| EQ-Bench Creative Writing | 1258 | — |
Frequently asked questions
Is GPT-4.5 better than Qwen2.5-Max?
Qwen2.5-Max is the stronger model overall, scoring 40.7 to 37.2 on the Noometry Index.
Is GPT-4.5 or Qwen2.5-Max better for coding?
They score almost the same on coding (42.2 vs 41.8); test both on your own repository before choosing.
How many benchmarks do GPT-4.5 and Qwen2.5-Max share?
26 benchmarks have published results for both models. GPT-4.5 has 42 scored results on Noometry and Qwen2.5-Max has 27.