Model comparison
Qwen2.5 72B Instruct vs Qwen3.6 Max Preview
Qwen3.6 Max Preview is the stronger model overall, scoring 51.5 to 31.9 on the Noometry Index.
Last verified . 19 shared benchmarks.
Alibaba (Qwen)Summary
- They share 19 benchmarks with published results for both. Qwen2.5 72B Instruct scores higher in 0 categories and Qwen3.6 Max Preview in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3.6 Max Preview leads 54.1 to 19.3.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 8.1% for Qwen2.5 72B Instruct and 91.1% for Qwen3.6 Max Preview.
- Qwen2.5 72B Instruct is cheaper at $1.40 / $5.60 per million input/output tokens, against $1.30 / $7.80 for Qwen3.6 Max Preview.
- Qwen3.6 Max Preview accepts more context: 262K tokens versus 131K.
- Qwen2.5 72B Instruct has downloadable open weights; the other is API-only.
Side by side
| Qwen2.5 72B Instruct | Qwen3.6 Max Preview | |
|---|---|---|
| Provider | Alibaba (Qwen) | Alibaba (Qwen) |
| Noometry Index | 31.9 | 51.5 |
| Released | 2024-09 | 2026-04-20 |
| Weights | Open | Proprietary |
| Context window | 131K | 262K |
| Max output | 8K | 66K |
| Input $ / M tokens | $1.40 | $1.30 |
| Output $ / M tokens | $5.60 | $7.80 |
| Results tracked | 43 | 29 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.6 Max Preview leads
Qwen2.5 72B Instruct: 33.2 (#260), Qwen3.6 Max Preview: 48.7 (#54)
| Benchmark | Qwen2.5 72B Instruct | Qwen3.6 Max Preview |
|---|---|---|
| LMArena Coding | 1292 | 1471 |
| SWE-bench Verified | — | 76.7% |
| LMArena WebDev | — | 1482 |
| WeirdML | 16% | — |
| BigCodeBench Instruct | 45.8% | — |
| BigCodeBench Complete | 55.9% | — |
Agentic & Tool Use Not comparable
Qwen2.5 72B Instruct: 22.1 (#133), Qwen3.6 Max Preview: —
| Benchmark | Qwen2.5 72B Instruct | Qwen3.6 Max Preview |
|---|---|---|
| TheAgentCompany | 5.7% | — |
| BALROG | 16.2% | — |
| METR Time Horizons | 35.8% | — |
| Vending-Bench 2 | — | 4,254 |
Reasoning Qwen3.6 Max Preview leads
Qwen2.5 72B Instruct: 22.3 (#199), Qwen3.6 Max Preview: 41.7 (#53)
| Benchmark | Qwen2.5 72B Instruct | Qwen3.6 Max Preview |
|---|---|---|
| LMArena Hard Prompts | 1271 | 1457 |
| DTBench | 62.9% | 87.2% |
| LMCA | 13.4% | 42.5% |
| Epoch Capabilities Index | 129 | 149.24 |
| SimpleBench | — | 63% |
| NYT Connections (extended) | — | 74.1% |
| Chess Puzzles | — | 20% |
| Mystery Game Puzzles | — | 19% |
| BIG-Bench Hard | 79.8% | — |
| ForecastBench | 57.5 | — |
| HellaSwag | 84.8% | — |
| PIQA | 82.6% | — |
| WinoGrande | 82.3% | — |
Math Qwen3.6 Max Preview leads
Qwen2.5 72B Instruct: 19.3 (#287), Qwen3.6 Max Preview: 54.1 (#46)
| Benchmark | Qwen2.5 72B Instruct | Qwen3.6 Max Preview |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 8.1% | 91.1% |
| LMArena Math | 1283 | 1465 |
| Omni-MATH | 33% | — |
| MATH Level 5 | 63.2% | — |
| FrontierMath (Feb 2025 set) | — | 23.1% |
| FrontierMath Tier 4 (v1) | — | 4.2% |
Knowledge Qwen3.6 Max Preview leads
Qwen2.5 72B Instruct: 27.0 (#253), Qwen3.6 Max Preview: 57.6 (#39)
| Benchmark | Qwen2.5 72B Instruct | Qwen3.6 Max Preview |
|---|---|---|
| GPQA Diamond | 49.1% | 87.4% |
| LMArena Expert | 1245 | 1478 |
| SimpleQA Verified | — | 52% |
| MMLU-Pro | 63.1% | — |
| Confabulations | 19.1% | — |
| GPQA (HELM) | 42.6% | — |
| ARC (AI2) Challenge | 94.5% | — |
| MMLU | 85.3% | — |
| TriviaQA | 71.9% | — |
Multilingual Qwen3.6 Max Preview leads
Qwen2.5 72B Instruct: 41.0 (#213), Qwen3.6 Max Preview: 54.2 (#48)
| Benchmark | Qwen2.5 72B Instruct | Qwen3.6 Max Preview |
|---|---|---|
| LMArena Non-English | 1252 | 1437 |
| LMArena Chinese | 1272 | 1487 |
| LMArena French | 1280 | 1449 |
| LMArena Russian | 1264 | 1445 |
| LMArena Spanish | 1256 | 1454 |
| LMArena German | 1234 | — |
| LMArena Japanese | 1180 | — |
| LMArena Korean | 1188 | — |
Instruction Following Qwen3.6 Max Preview leads
Qwen2.5 72B Instruct: 65.5 (#221), Qwen3.6 Max Preview: 75.7 (#55)
| Benchmark | Qwen2.5 72B Instruct | Qwen3.6 Max Preview |
|---|---|---|
| LMArena Instruction Following | 1254 | 1438 |
| IFEval | 80.6% | — |
Long Context Qwen3.6 Max Preview leads
Qwen2.5 72B Instruct: 38.9 (#188), Qwen3.6 Max Preview: 44.6 (#61)
| Benchmark | Qwen2.5 72B Instruct | Qwen3.6 Max Preview |
|---|---|---|
| LMArena Longer Query | 1282 | 1457 |
Writing & Preference Qwen3.6 Max Preview leads
Qwen2.5 72B Instruct: 46.7 (#215), Qwen3.6 Max Preview: 63.8 (#60)
| Benchmark | Qwen2.5 72B Instruct | Qwen3.6 Max Preview |
|---|---|---|
| LMArena Text | 1269 | 1447 |
| LMArena Creative Writing | 1221 | 1435 |
| LMArena Multi-Turn | 1272 | 1456 |
| WildBench | 80.2% | — |
Frequently asked questions
Is Qwen2.5 72B Instruct better than Qwen3.6 Max Preview?
Qwen3.6 Max Preview is the stronger model overall, scoring 51.5 to 31.9 on the Noometry Index.
Which is cheaper, Qwen2.5 72B Instruct or Qwen3.6 Max Preview?
Qwen2.5 72B Instruct is cheaper. It lists at $1.40 per million input tokens and $5.60 per million output tokens; Qwen3.6 Max Preview lists at $1.30 and $7.80.
Is Qwen2.5 72B Instruct or Qwen3.6 Max Preview better for coding?
Qwen3.6 Max Preview scores higher on coding benchmarks: 48.7 versus 33.2 in the Noometry coding category.
Which has the bigger context window?
Qwen3.6 Max Preview does, with 262K tokens against 131K.
How many benchmarks do Qwen2.5 72B Instruct and Qwen3.6 Max Preview share?
19 benchmarks have published results for both models. Qwen2.5 72B Instruct has 43 scored results on Noometry and Qwen3.6 Max Preview has 29.