Model comparison
DeepSeek-V3 vs Qwen1.5-32B
DeepSeek-V3 is the stronger model overall, scoring 39.5 to 30.5 on the Noometry Index.
Last verified . 21 shared benchmarks.
DeepSeek
Alibaba (Qwen)Summary
- They share 21 benchmarks with published results for both. DeepSeek-V3 scores higher in 5 categories and Qwen1.5-32B in 3 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where DeepSeek-V3 leads 37.5 to 13.5.
- The biggest single-benchmark swing is GPQA Diamond: 67.6% for DeepSeek-V3 and 30.7% for Qwen1.5-32B.
Side by side
| DeepSeek-V3 | Qwen1.5-32B | |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 39.5 | 30.5 |
| Released | 2024-12-26 | 2024-02-04 |
| Weights | Open | Open |
| Context window | 164K | — |
| Max output | 164K | — |
| Input $ / M tokens | $0.24 | — |
| Output $ / M tokens | $0.90 | — |
| Results tracked | 60 | 21 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-V3 leads
DeepSeek-V3: 42.3 (#106), Qwen1.5-32B: 31.7 (#282)
| Benchmark | DeepSeek-V3 | Qwen1.5-32B |
|---|---|---|
| BigCodeBench Instruct | 50% | 32.3% |
| LMArena Coding | 1368 | 1155 |
| BigCodeBench Complete | 62.2% | 42% |
| Aider Polyglot | 55.1% | — |
| SciCode | 35.8% | — |
| WeirdML | 36.1% | — |
| LiveBench Coding | 70.9% | — |
| HumanEval+ | 86.6% | — |
| MBPP+ | 73% | — |
Agentic & Tool Use Not comparable
DeepSeek-V3: —, Qwen1.5-32B: —
| Benchmark | DeepSeek-V3 | Qwen1.5-32B |
|---|---|---|
| METR Time Horizons | 49.6% | — |
Reasoning Qwen1.5-32B leads
DeepSeek-V3: 20.5 (#236), Qwen1.5-32B: 21.8 (#212)
| Benchmark | DeepSeek-V3 | Qwen1.5-32B |
|---|---|---|
| LMArena Hard Prompts | 1365 | 1130 |
| SimpleBench | 27.2% | — |
| Kagi LLM Benchmark | 52.3% | — |
| CritPt | 0% | — |
| LiveBench Reasoning | 65.8% | — |
| DTBench | 64.8% | — |
| LiveBench Data Analysis | 60.9% | — |
| LMCA | 15.5% | — |
| BIG-Bench Hard | 87.5% | — |
| Epoch Capabilities Index | 135.94 | — |
| ForecastBench | 59.1 | — |
| HellaSwag | 88.9% | — |
| LiveBench | 66.9% | — |
| PIQA | 84.7% | — |
| WinoGrande | 85.2% | — |
Math Too close to call
DeepSeek-V3: 32.1 (#219), Qwen1.5-32B: 33.0 (#207)
| Benchmark | DeepSeek-V3 | Qwen1.5-32B |
|---|---|---|
| LMArena Math | 1373 | 1155 |
| OTIS Mock AIME 2024-2025 | 37.8% | — |
| Omni-MATH | 40.3% | — |
| LiveBench Math | 73.5% | — |
| MATH Level 5 | 75.5% | — |
| FrontierMath (Feb 2025 set) | 1.7% | — |
Knowledge DeepSeek-V3 leads
DeepSeek-V3: 37.5 (#155), Qwen1.5-32B: 13.5 (#296)
| Benchmark | DeepSeek-V3 | Qwen1.5-32B |
|---|---|---|
| GPQA Diamond | 67.6% | 30.7% |
| LMArena Expert | 1351 | 1126 |
| MMLU | 87.2% | 74.4% |
| MMLU-Pro | 72.3% | — |
| Confabulations | 26.1% | — |
| Vectara Hallucination Rate | 6.1% | — |
| GPQA (HELM) | 53.8% | — |
| ARC (AI2) Challenge | 95.3% | — |
| TriviaQA | 82.9% | — |
Multilingual DeepSeek-V3 leads
DeepSeek-V3: 48.5 (#143), Qwen1.5-32B: 31.4 (#259)
| Benchmark | DeepSeek-V3 | Qwen1.5-32B |
|---|---|---|
| LMArena Non-English | 1358 | 1106 |
| LMArena Chinese | 1391 | 1177 |
| LMArena French | 1385 | 1101 |
| LMArena German | 1374 | 1058 |
| LMArena Japanese | 1333 | 1027 |
| LMArena Korean | 1319 | 1008 |
| LMArena Russian | 1373 | 1073 |
| LMArena Spanish | 1358 | 1089 |
Instruction Following DeepSeek-V3 leads
DeepSeek-V3: 72.8 (#130), Qwen1.5-32B: 57.7 (#265)
| Benchmark | DeepSeek-V3 | Qwen1.5-32B |
|---|---|---|
| LMArena Instruction Following | 1345 | 1116 |
| LiveBench Instruction Following | 81.5% | — |
| IFEval | 83.2% | — |
Long Context Too close to call
DeepSeek-V3: 34.0 (#253), Qwen1.5-32B: 34.7 (#246)
| Benchmark | DeepSeek-V3 | Qwen1.5-32B |
|---|---|---|
| LMArena Longer Query | 1352 | 1146 |
| Fiction.LiveBench | 50% | — |
Writing & Preference DeepSeek-V3 leads
DeepSeek-V3: 57.4 (#130), Qwen1.5-32B: 34.2 (#271)
| Benchmark | DeepSeek-V3 | Qwen1.5-32B |
|---|---|---|
| LMArena Text | 1375 | 1137 |
| LMArena Creative Writing | 1364 | 1083 |
| LMArena Multi-Turn | 1389 | 1140 |
| Short-Story Creative Writing | 77% | — |
| EQ-Bench Creative Writing | 1472 | — |
| WildBench | 83% | — |
| LiveBench Language | 49.1% | — |
Frequently asked questions
Is DeepSeek-V3 better than Qwen1.5-32B?
DeepSeek-V3 is the stronger model overall, scoring 39.5 to 30.5 on the Noometry Index.
Is DeepSeek-V3 or Qwen1.5-32B better for coding?
DeepSeek-V3 scores higher on coding benchmarks: 42.3 versus 31.7 in the Noometry coding category.
How many benchmarks do DeepSeek-V3 and Qwen1.5-32B share?
21 benchmarks have published results for both models. DeepSeek-V3 has 60 scored results on Noometry and Qwen1.5-32B has 21.