Model comparison
Qwen1.5-72B vs Qwen3 32B
Qwen3 32B is the stronger model overall, scoring 38.8 to 30.3 on the Noometry Index.
Last verified . 14 shared benchmarks.
Alibaba (Qwen)Summary
- They share 14 benchmarks with published results for both. Qwen1.5-72B scores higher in 1 category and Qwen3 32B in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Qwen3 32B leads 40.0 to 11.5.
- The biggest single-benchmark swing is GPQA Diamond: 28.8% for Qwen1.5-72B and 65.7% for Qwen3 32B.
Side by side
| Qwen1.5-72B | Qwen3 32B | |
|---|---|---|
| Provider | Alibaba (Qwen) | Alibaba (Qwen) |
| Noometry Index | 30.3 | 38.8 |
| Released | 2024-02-04 | 2025-04 |
| Weights | Open | Open |
| Context window | — | 131K |
| Max output | — | 16K |
| Input $ / M tokens | — | $0.70 |
| Output $ / M tokens | — | $2.80 |
| Results tracked | 22 | 26 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3 32B leads
Qwen1.5-72B: 31.4 (#272), Qwen3 32B: 37.3 (#184)
| Benchmark | Qwen1.5-72B | Qwen3 32B |
|---|---|---|
| LMArena Coding | 1166 | 1358 |
| Aider Polyglot | — | 40% |
| SciCode | — | 35.4% |
| BigCodeBench Instruct | 33.2% | — |
| BigCodeBench Complete | 40.3% | — |
| HumanEval+ | 59.1% | — |
| MBPP+ | 61.6% | — |
Agentic & Tool Use Not comparable
Qwen1.5-72B: —, Qwen3 32B: 32.3 (#58)
| Benchmark | Qwen1.5-72B | Qwen3 32B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 48.7% |
Reasoning Qwen1.5-72B leads
Qwen1.5-72B: 21.2 (#201), Qwen3 32B: 19.6 (#233)
| Benchmark | Qwen1.5-72B | Qwen3 32B |
|---|---|---|
| LMArena Hard Prompts | 1148 | 1335 |
| Kagi LLM Benchmark | — | 54.9% |
| CritPt | — | 0.3% |
| Chess Puzzles | — | 5% |
| DTBench | — | 67.5% |
| LMCA | — | 17.3% |
| Epoch Capabilities Index | — | 138.51 |
Math Qwen3 32B leads
Qwen1.5-72B: 30.7 (#205), Qwen3 32B: 37.9 (#89)
| Benchmark | Qwen1.5-72B | Qwen3 32B |
|---|---|---|
| LMArena Math | 1164 | 1399 |
| OTIS Mock AIME 2024-2025 | — | 66.9% |
Knowledge Qwen3 32B leads
Qwen1.5-72B: 11.5 (#295), Qwen3 32B: 40.0 (#112)
| Benchmark | Qwen1.5-72B | Qwen3 32B |
|---|---|---|
| GPQA Diamond | 28.8% | 65.7% |
| LMArena Expert | 1136 | 1362 |
| Vectara Hallucination Rate | — | 5.9% |
Multilingual Qwen3 32B leads
Qwen1.5-72B: 33.4 (#248), Qwen3 32B: 45.8 (#162)
| Benchmark | Qwen1.5-72B | Qwen3 32B |
|---|---|---|
| LMArena Non-English | 1135 | 1317 |
| LMArena Chinese | 1186 | 1357 |
| LMArena German | 1084 | 1341 |
| LMArena Russian | 1104 | 1311 |
| LMArena French | 1158 | — |
| LMArena Japanese | 1061 | — |
| LMArena Korean | 1049 | — |
| LMArena Spanish | 1110 | — |
Instruction Following Qwen3 32B leads
Qwen1.5-72B: 59.6 (#251), Qwen3 32B: 69.1 (#174)
| Benchmark | Qwen1.5-72B | Qwen3 32B |
|---|---|---|
| LMArena Instruction Following | 1142 | 1305 |
Long Context Qwen3 32B leads
Qwen1.5-72B: 35.0 (#238), Qwen3 32B: 43.7 (#83)
| Benchmark | Qwen1.5-72B | Qwen3 32B |
|---|---|---|
| LMArena Longer Query | 1157 | 1327 |
| Fiction.LiveBench | — | 74.2% |
Writing & Preference Qwen3 32B leads
Qwen1.5-72B: 37.4 (#253), Qwen3 32B: 53.1 (#158)
| Benchmark | Qwen1.5-72B | Qwen3 32B |
|---|---|---|
| LMArena Text | 1166 | 1340 |
| LMArena Creative Writing | 1137 | 1297 |
| LMArena Multi-Turn | 1160 | 1331 |
Frequently asked questions
Is Qwen1.5-72B better than Qwen3 32B?
Qwen3 32B is the stronger model overall, scoring 38.8 to 30.3 on the Noometry Index.
Is Qwen1.5-72B or Qwen3 32B better for coding?
Qwen3 32B scores higher on coding benchmarks: 37.3 versus 31.4 in the Noometry coding category.
How many benchmarks do Qwen1.5-72B and Qwen3 32B share?
14 benchmarks have published results for both models. Qwen1.5-72B has 22 scored results on Noometry and Qwen3 32B has 26.