Model comparison
Qwen2.5 72B Instruct vs Qwen3.6 27B
Qwen3.6 27B is the stronger model overall, scoring 42.2 to 31.9 on the Noometry Index.
Last verified . 5 shared benchmarks.
Alibaba (Qwen)Summary
- They share 5 benchmarks with published results for both. Qwen2.5 72B Instruct scores higher in 0 categories and Qwen3.6 27B in 5 categories; 5 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3.6 27B leads 48.5 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 27B.
- Qwen3.6 27B is cheaper at $0.60 / $3.60 per million input/output tokens, against $1.40 / $5.60 for Qwen2.5 72B Instruct.
- Qwen3.6 27B accepts more context: 262K tokens versus 131K.
Side by side
| Qwen2.5 72B Instruct | Qwen3.6 27B | |
|---|---|---|
| Provider | Alibaba (Qwen) | Alibaba (Qwen) |
| Noometry Index | 31.9 | 42.2 |
| Released | 2024-09 | 2026-04-22 |
| Weights | Open | Open |
| Context window | 131K | 262K |
| Max output | 8K | 66K |
| Input $ / M tokens | $1.40 | $0.60 |
| Output $ / M tokens | $5.60 | $3.60 |
| Results tracked | 43 | 11 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.6 27B leads
Qwen2.5 72B Instruct: 33.2 (#260), Qwen3.6 27B: 39.1 (#163)
| Benchmark | Qwen2.5 72B Instruct | Qwen3.6 27B |
|---|---|---|
| SciCode | — | 37.3% |
| WeirdML | 16% | — |
| BigCodeBench Instruct | 45.8% | — |
| LMArena Coding | 1292 | — |
| BigCodeBench Complete | 55.9% | — |
Agentic & Tool Use Not comparable
Qwen2.5 72B Instruct: 22.1 (#133), Qwen3.6 27B: —
| Benchmark | Qwen2.5 72B Instruct | Qwen3.6 27B |
|---|---|---|
| TheAgentCompany | 5.7% | — |
| BALROG | 16.2% | — |
| METR Time Horizons | 35.8% | — |
Reasoning Qwen3.6 27B leads
Qwen2.5 72B Instruct: 22.3 (#199), Qwen3.6 27B: 25.0 (#153)
| Benchmark | Qwen2.5 72B Instruct | Qwen3.6 27B |
|---|---|---|
| DTBench | 62.9% | 78.1% |
| LMCA | 13.4% | 34.5% |
| Epoch Capabilities Index | 129 | 146.5 |
| CritPt | — | 0.9% |
| Chess Puzzles | — | 22% |
| LMArena Hard Prompts | 1271 | — |
| Mystery Game Puzzles | — | 7% |
| BIG-Bench Hard | 79.8% | — |
| ForecastBench | 57.5 | — |
| HellaSwag | 84.8% | — |
| PIQA | 82.6% | — |
| WinoGrande | 82.3% | — |
Math Qwen3.6 27B leads
Qwen2.5 72B Instruct: 19.3 (#287), Qwen3.6 27B: 48.5 (#62)
| Benchmark | Qwen2.5 72B Instruct | Qwen3.6 27B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 8.1% | 91.1% |
| FrontierMath (Tiers 1-3) | — | 35.1% |
| Omni-MATH | 33% | — |
| LMArena Math | 1283 | — |
| MATH Level 5 | 63.2% | — |
Knowledge Qwen3.6 27B leads
Qwen2.5 72B Instruct: 27.0 (#253), Qwen3.6 27B: 52.4 (#63)
| Benchmark | Qwen2.5 72B Instruct | Qwen3.6 27B |
|---|---|---|
| GPQA Diamond | 49.1% | 85.9% |
| MMLU-Pro | 63.1% | — |
| Confabulations | 19.1% | — |
| GPQA (HELM) | 42.6% | — |
| LMArena Expert | 1245 | — |
| ARC (AI2) Challenge | 94.5% | — |
| MMLU | 85.3% | — |
| TriviaQA | 71.9% | — |
Multilingual Not comparable
Qwen2.5 72B Instruct: 41.0 (#213), Qwen3.6 27B: —
| Benchmark | Qwen2.5 72B Instruct | Qwen3.6 27B |
|---|---|---|
| LMArena Non-English | 1252 | — |
| LMArena Chinese | 1272 | — |
| LMArena French | 1280 | — |
| LMArena German | 1234 | — |
| LMArena Japanese | 1180 | — |
| LMArena Korean | 1188 | — |
| LMArena Russian | 1264 | — |
| LMArena Spanish | 1256 | — |
Instruction Following Not comparable
Qwen2.5 72B Instruct: 65.5 (#221), Qwen3.6 27B: —
| Benchmark | Qwen2.5 72B Instruct | Qwen3.6 27B |
|---|---|---|
| IFEval | 80.6% | — |
| LMArena Instruction Following | 1254 | — |
Long Context Not comparable
Qwen2.5 72B Instruct: 38.9 (#188), Qwen3.6 27B: —
| Benchmark | Qwen2.5 72B Instruct | Qwen3.6 27B |
|---|---|---|
| LMArena Longer Query | 1282 | — |
Writing & Preference Qwen3.6 27B leads
Qwen2.5 72B Instruct: 46.7 (#215), Qwen3.6 27B: 50.3 (#181)
| Benchmark | Qwen2.5 72B Instruct | Qwen3.6 27B |
|---|---|---|
| LMArena Text | 1269 | — |
| LMArena Creative Writing | 1221 | — |
| WildBench | 80.2% | — |
| EQ-Bench 4 | — | 1026 |
| LMArena Multi-Turn | 1272 | — |
Frequently asked questions
Is Qwen2.5 72B Instruct better than Qwen3.6 27B?
Qwen3.6 27B is the stronger model overall, scoring 42.2 to 31.9 on the Noometry Index.
Which is cheaper, Qwen2.5 72B Instruct or Qwen3.6 27B?
Qwen3.6 27B is cheaper. It lists at $0.60 per million input tokens and $3.60 per million output tokens; Qwen2.5 72B Instruct lists at $1.40 and $5.60.
Is Qwen2.5 72B Instruct or Qwen3.6 27B better for coding?
Qwen3.6 27B scores higher on coding benchmarks: 39.1 versus 33.2 in the Noometry coding category.
Which has the bigger context window?
Qwen3.6 27B does, with 262K tokens against 131K.
How many benchmarks do Qwen2.5 72B Instruct and Qwen3.6 27B share?
5 benchmarks have published results for both models. Qwen2.5 72B Instruct has 43 scored results on Noometry and Qwen3.6 27B has 11.