Model comparison
Qwen2-72B vs Qwen3.5 35B-A3B
Qwen3.5 35B-A3B is the stronger model overall, scoring 42.0 to 30.0 on the Noometry Index.
Last verified . 19 shared benchmarks.
Alibaba (Qwen)Summary
- They share 19 benchmarks with published results for both. Qwen2-72B scores higher in 0 categories and Qwen3.5 35B-A3B in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Qwen3.5 35B-A3B leads 47.8 to 21.2.
- The biggest single-benchmark swing is GPQA Diamond: 40.8% for Qwen2-72B and 83.5% for Qwen3.5 35B-A3B.
Side by side
| Qwen2-72B | Qwen3.5 35B-A3B | |
|---|---|---|
| Provider | Alibaba (Qwen) | Alibaba (Qwen) |
| Noometry Index | 30.0 | 42.0 |
| Released | 2024-06-07 | 2026-02-01 |
| Weights | Open | Open |
| Context window | — | 262K |
| Max output | — | 66K |
| Input $ / M tokens | — | $0.25 |
| Output $ / M tokens | — | $2 |
| Results tracked | 26 | 28 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.5 35B-A3B leads
Qwen2-72B: 29.1 (#310), Qwen3.5 35B-A3B: 33.8 (#251)
| Benchmark | Qwen2-72B | Qwen3.5 35B-A3B |
|---|---|---|
| LMArena Coding | 1196 | 1410 |
| LMArena WebDev | — | 1254 |
| SciCode | — | 29.3% |
| WeirdML | 11.3% | — |
| BigCodeBench Instruct | 38.5% | — |
| BigCodeBench Complete | 54% | — |
Agentic & Tool Use Not comparable
Qwen2-72B: 17.0 (#146), Qwen3.5 35B-A3B: —
| Benchmark | Qwen2-72B | Qwen3.5 35B-A3B |
|---|---|---|
| TheAgentCompany | 1.1% | — |
| METR Time Horizons | 29.9% | — |
Reasoning Qwen3.5 35B-A3B leads
Qwen2-72B: 23.2 (#181), Qwen3.5 35B-A3B: 24.6 (#161)
| Benchmark | Qwen2-72B | Qwen3.5 35B-A3B |
|---|---|---|
| LMArena Hard Prompts | 1191 | 1400 |
| Epoch Capabilities Index | 125.28 | 142.52 |
| CritPt | — | 0.6% |
| Chess Puzzles | — | 10% |
| DTBench | — | 80% |
| LMCA | — | 29.5% |
Math Qwen3.5 35B-A3B leads
Qwen2-72B: 30.2 (#236), Qwen3.5 35B-A3B: 39.9 (#97)
| Benchmark | Qwen2-72B | Qwen3.5 35B-A3B |
|---|---|---|
| LMArena Math | 1235 | 1404 |
| MathArena Final-Answer Competitions | — | 56% |
| OTIS Mock AIME 2024-2025 | — | 70% |
| MATH Level 5 | 39.1% | — |
Knowledge Qwen3.5 35B-A3B leads
Qwen2-72B: 21.2 (#275), Qwen3.5 35B-A3B: 47.8 (#79)
| Benchmark | Qwen2-72B | Qwen3.5 35B-A3B |
|---|---|---|
| GPQA Diamond | 40.8% | 83.5% |
| LMArena Expert | 1171 | 1408 |
| Vectara Hallucination Rate | — | 10.5% |
| MMLU | 82.4% | — |
Multilingual Qwen3.5 35B-A3B leads
Qwen2-72B: 35.9 (#244), Qwen3.5 35B-A3B: 50.0 (#127)
| Benchmark | Qwen2-72B | Qwen3.5 35B-A3B |
|---|---|---|
| LMArena Non-English | 1176 | 1378 |
| LMArena Chinese | 1240 | 1457 |
| LMArena French | 1170 | 1412 |
| LMArena German | 1151 | 1367 |
| LMArena Japanese | 1111 | 1325 |
| LMArena Korean | 1083 | 1356 |
| LMArena Russian | 1169 | 1376 |
| LMArena Spanish | 1169 | 1392 |
Instruction Following Qwen3.5 35B-A3B leads
Qwen2-72B: 61.7 (#241), Qwen3.5 35B-A3B: 72.8 (#128)
| Benchmark | Qwen2-72B | Qwen3.5 35B-A3B |
|---|---|---|
| LMArena Instruction Following | 1181 | 1379 |
Long Context Qwen3.5 35B-A3B leads
Qwen2-72B: 36.1 (#235), Qwen3.5 35B-A3B: 42.4 (#127)
| Benchmark | Qwen2-72B | Qwen3.5 35B-A3B |
|---|---|---|
| LMArena Longer Query | 1192 | 1389 |
Writing & Preference Qwen3.5 35B-A3B leads
Qwen2-72B: 40.8 (#241), Qwen3.5 35B-A3B: 57.9 (#124)
| Benchmark | Qwen2-72B | Qwen3.5 35B-A3B |
|---|---|---|
| LMArena Text | 1203 | 1395 |
| LMArena Creative Writing | 1181 | 1346 |
| LMArena Multi-Turn | 1196 | 1390 |
Frequently asked questions
Is Qwen2-72B better than Qwen3.5 35B-A3B?
Qwen3.5 35B-A3B is the stronger model overall, scoring 42.0 to 30.0 on the Noometry Index.
Is Qwen2-72B or Qwen3.5 35B-A3B better for coding?
Qwen3.5 35B-A3B scores higher on coding benchmarks: 33.8 versus 29.1 in the Noometry coding category.
How many benchmarks do Qwen2-72B and Qwen3.5 35B-A3B share?
19 benchmarks have published results for both models. Qwen2-72B has 26 scored results on Noometry and Qwen3.5 35B-A3B has 28.