Model comparison
Qwen Max vs Qwen2.5 72B Instruct
Qwen Max is the stronger model overall, scoring 34.7 to 31.9 on the Noometry Index.
Last verified . 20 shared benchmarks.
Alibaba (Qwen)Summary
- They share 20 benchmarks with published results for both. Qwen Max scores higher in 7 categories and Qwen2.5 72B Instruct in 1 category; 6 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Qwen Max leads 30.3 to 27.0.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 16.1% for Qwen Max and 8.1% for Qwen2.5 72B Instruct.
- Qwen2.5 72B Instruct is cheaper at $1.40 / $5.60 per million input/output tokens, against $1.60 / $6.40 for Qwen Max.
- Qwen2.5 72B Instruct accepts more context: 131K tokens versus 33K.
- Qwen2.5 72B Instruct has downloadable open weights; the other is API-only.
Side by side
| Qwen Max | Qwen2.5 72B Instruct | |
|---|---|---|
| Provider | Alibaba (Qwen) | Alibaba (Qwen) |
| Noometry Index | 34.7 | 31.9 |
| Released | 2024-04-03 | 2024-09 |
| Weights | Proprietary | Open |
| Context window | 33K | 131K |
| Max output | 8K | 8K |
| Input $ / M tokens | $1.60 | $1.40 |
| Output $ / M tokens | $6.40 | $5.60 |
| Results tracked | 23 | 43 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen2.5 72B Instruct leads
Qwen Max: 30.7 (#292), Qwen2.5 72B Instruct: 33.2 (#260)
| Benchmark | Qwen Max | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Coding | 1288 | 1292 |
| Aider Polyglot | 21.8% | — |
| WeirdML | — | 16% |
| BigCodeBench Instruct | — | 45.8% |
| BigCodeBench Complete | — | 55.9% |
Agentic & Tool Use Not comparable
Qwen Max: —, Qwen2.5 72B Instruct: 22.1 (#133)
| Benchmark | Qwen Max | Qwen2.5 72B Instruct |
|---|---|---|
| TheAgentCompany | — | 5.7% |
| BALROG | — | 16.2% |
| METR Time Horizons | — | 35.8% |
Reasoning Qwen Max leads
Qwen Max: 25.1 (#151), Qwen2.5 72B Instruct: 22.3 (#199)
| Benchmark | Qwen Max | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Hard Prompts | 1269 | 1271 |
| DTBench | — | 62.9% |
| LMCA | — | 13.4% |
| BIG-Bench Hard | — | 79.8% |
| Epoch Capabilities Index | — | 129 |
| ForecastBench | — | 57.5 |
| HellaSwag | — | 84.8% |
| PIQA | — | 82.6% |
| WinoGrande | — | 82.3% |
Math Qwen Max leads
Qwen Max: 22.3 (#276), Qwen2.5 72B Instruct: 19.3 (#287)
| Benchmark | Qwen Max | Qwen2.5 72B Instruct |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 16.1% | 8.1% |
| LMArena Math | 1275 | 1283 |
| MATH Level 5 | 67.2% | 63.2% |
| Omni-MATH | — | 33% |
| FrontierMath (Feb 2025 set) | 1% | — |
Knowledge Qwen Max leads
Qwen Max: 30.3 (#228), Qwen2.5 72B Instruct: 27.0 (#253)
| Benchmark | Qwen Max | Qwen2.5 72B Instruct |
|---|---|---|
| GPQA Diamond | 56.1% | 49.1% |
| LMArena Expert | 1248 | 1245 |
| MMLU-Pro | — | 63.1% |
| Confabulations | — | 19.1% |
| GPQA (HELM) | — | 42.6% |
| ARC (AI2) Challenge | — | 94.5% |
| MMLU | — | 85.3% |
| TriviaQA | — | 71.9% |
Multilingual Too close to call
Qwen Max: 41.8 (#202), Qwen2.5 72B Instruct: 41.0 (#213)
| Benchmark | Qwen Max | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Non-English | 1263 | 1252 |
| LMArena Chinese | 1254 | 1272 |
| LMArena French | 1330 | 1280 |
| LMArena German | 1254 | 1234 |
| LMArena Japanese | 1205 | 1180 |
| LMArena Korean | 1142 | 1188 |
| LMArena Russian | 1274 | 1264 |
| LMArena Spanish | 1290 | 1256 |
Instruction Following Qwen Max leads
Qwen Max: 66.5 (#208), Qwen2.5 72B Instruct: 65.5 (#221)
| Benchmark | Qwen Max | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Instruction Following | 1262 | 1254 |
| IFEval | — | 80.6% |
Long Context Too close to call
Qwen Max: 39.4 (#180), Qwen2.5 72B Instruct: 38.9 (#188)
| Benchmark | Qwen Max | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Longer Query | 1288 | 1282 |
| Fiction.LiveBench | 66.7% | — |
Writing & Preference Qwen Max leads
Qwen Max: 47.8 (#205), Qwen2.5 72B Instruct: 46.7 (#215)
| Benchmark | Qwen Max | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Text | 1282 | 1269 |
| LMArena Creative Writing | 1248 | 1221 |
| LMArena Multi-Turn | 1277 | 1272 |
| WildBench | — | 80.2% |
Frequently asked questions
Is Qwen Max better than Qwen2.5 72B Instruct?
Qwen Max is the stronger model overall, scoring 34.7 to 31.9 on the Noometry Index.
Which is cheaper, Qwen Max or Qwen2.5 72B Instruct?
Qwen2.5 72B Instruct is cheaper. It lists at $1.40 per million input tokens and $5.60 per million output tokens; Qwen Max lists at $1.60 and $6.40.
Is Qwen Max or Qwen2.5 72B Instruct better for coding?
Qwen2.5 72B Instruct scores higher on coding benchmarks: 33.2 versus 30.7 in the Noometry coding category.
Which has the bigger context window?
Qwen2.5 72B Instruct does, with 131K tokens against 33K.
How many benchmarks do Qwen Max and Qwen2.5 72B Instruct share?
20 benchmarks have published results for both models. Qwen Max has 23 scored results on Noometry and Qwen2.5 72B Instruct has 43.