Model comparison
Qwen2-72B vs Qwen3 Max
Qwen3 Max is the stronger model overall, scoring 43.7 to 30.0 on the Noometry Index.
Last verified . 20 shared benchmarks.
Alibaba (Qwen)Summary
- They share 20 benchmarks with published results for both. Qwen2-72B scores higher in 1 category and Qwen3 Max in 7 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Qwen3 Max leads 48.1 to 21.2.
- The biggest single-benchmark swing is MATH Level 5: 39.1% for Qwen2-72B and 97.1% for Qwen3 Max.
- Qwen2-72B has downloadable open weights; the other is API-only.
Side by side
| Qwen2-72B | Qwen3 Max | |
|---|---|---|
| Provider | Alibaba (Qwen) | Alibaba (Qwen) |
| Noometry Index | 30.0 | 43.7 |
| Released | 2024-06-07 | 2025-09-23 |
| Weights | Open | Proprietary |
| Context window | — | 262K |
| Max output | — | 66K |
| Input $ / M tokens | — | $1.20 |
| Output $ / M tokens | — | $6 |
| Results tracked | 26 | 33 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3 Max leads
Qwen2-72B: 29.1 (#310), Qwen3 Max: 43.0 (#93)
| Benchmark | Qwen2-72B | Qwen3 Max |
|---|---|---|
| LMArena Coding | 1196 | 1456 |
| WeirdML | 11.3% | — |
| BigCodeBench Instruct | 38.5% | — |
| BigCodeBench Complete | 54% | — |
| ALE-Bench | — | 370.45 |
Agentic & Tool Use Not comparable
Qwen2-72B: 17.0 (#146), Qwen3 Max: —
| Benchmark | Qwen2-72B | Qwen3 Max |
|---|---|---|
| TheAgentCompany | 1.1% | — |
| METR Time Horizons | 29.9% | — |
| Vending-Bench 2 | — | 71.56 |
Reasoning Too close to call
Qwen2-72B: 23.2 (#181), Qwen3 Max: 22.6 (#190)
| Benchmark | Qwen2-72B | Qwen3 Max |
|---|---|---|
| LMArena Hard Prompts | 1191 | 1448 |
| Epoch Capabilities Index | 125.28 | 142.38 |
| Kagi LLM Benchmark | — | 72.5% |
| NYT Connections (extended) | — | 30.1% |
| Chess Puzzles | — | 4% |
| Mystery Game Puzzles | — | 5% |
| DTBench | — | 82.1% |
| LMCA | — | 28.3% |
Math Qwen3 Max leads
Qwen2-72B: 30.2 (#236), Qwen3 Max: 38.7 (#131)
| Benchmark | Qwen2-72B | Qwen3 Max |
|---|---|---|
| LMArena Math | 1235 | 1446 |
| MATH Level 5 | 39.1% | 97.1% |
| FrontierMath (Tiers 1-3) | — | 18.9% |
| OTIS Mock AIME 2024-2025 | — | 73.3% |
Knowledge Qwen3 Max leads
Qwen2-72B: 21.2 (#275), Qwen3 Max: 48.1 (#78)
| Benchmark | Qwen2-72B | Qwen3 Max |
|---|---|---|
| GPQA Diamond | 40.8% | 72.6% |
| LMArena Expert | 1171 | 1455 |
| SimpleQA Verified | — | 48.7% |
| MMLU | 82.4% | — |
Multilingual Qwen3 Max leads
Qwen2-72B: 35.9 (#244), Qwen3 Max: 53.7 (#62)
| Benchmark | Qwen2-72B | Qwen3 Max |
|---|---|---|
| LMArena Non-English | 1176 | 1429 |
| LMArena Chinese | 1240 | 1478 |
| LMArena French | 1170 | 1449 |
| LMArena German | 1151 | 1463 |
| LMArena Japanese | 1111 | 1397 |
| LMArena Korean | 1083 | 1399 |
| LMArena Russian | 1169 | 1428 |
| LMArena Spanish | 1169 | 1462 |
Instruction Following Qwen3 Max leads
Qwen2-72B: 61.7 (#241), Qwen3 Max: 74.8 (#87)
| Benchmark | Qwen2-72B | Qwen3 Max |
|---|---|---|
| LMArena Instruction Following | 1181 | 1419 |
Long Context Qwen3 Max leads
Qwen2-72B: 36.1 (#235), Qwen3 Max: 41.6 (#134)
| Benchmark | Qwen2-72B | Qwen3 Max |
|---|---|---|
| LMArena Longer Query | 1192 | 1438 |
| Fiction.LiveBench | — | 66.7% |
| CL-bench | — | 14.5% |
Writing & Preference Qwen3 Max leads
Qwen2-72B: 40.8 (#241), Qwen3 Max: 62.4 (#76)
| Benchmark | Qwen2-72B | Qwen3 Max |
|---|---|---|
| LMArena Text | 1203 | 1439 |
| LMArena Creative Writing | 1181 | 1402 |
| LMArena Multi-Turn | 1196 | 1446 |
Frequently asked questions
Is Qwen2-72B better than Qwen3 Max?
Qwen3 Max is the stronger model overall, scoring 43.7 to 30.0 on the Noometry Index.
Is Qwen2-72B or Qwen3 Max better for coding?
Qwen3 Max scores higher on coding benchmarks: 43.0 versus 29.1 in the Noometry coding category.
How many benchmarks do Qwen2-72B and Qwen3 Max share?
20 benchmarks have published results for both models. Qwen2-72B has 26 scored results on Noometry and Qwen3 Max has 33.