Model comparison
Deepseek Coder v2 vs QwQ-32B
QwQ-32B is the stronger model overall, scoring 39.8 to 35.9 on the Noometry Index.
Last verified . 19 shared benchmarks.
DeepSeek
Alibaba (Qwen)Summary
- They share 19 benchmarks with published results for both. Deepseek Coder v2 scores higher in 1 category and QwQ-32B in 7 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where QwQ-32B leads 50.6 to 38.2.
- The biggest single-benchmark swing is BigCodeBench Complete: 59.7% for Deepseek Coder v2 and 54.4% for QwQ-32B.
Side by side
| Deepseek Coder v2 | QwQ-32B | |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 35.9 | 39.8 |
| Released | 2024-06-17 | 2024-11-28 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 24 | 36 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Deepseek Coder v2 leads
Deepseek Coder v2: 38.1 (#183), QwQ-32B: 35.4 (#226)
| Benchmark | Deepseek Coder v2 | QwQ-32B |
|---|---|---|
| BigCodeBench Instruct | 48.2% | 44.6% |
| LMArena Coding | 1251 | 1333 |
| BigCodeBench Complete | 59.7% | 54.4% |
| Aider Polyglot | — | 20.9% |
| LiveBench Coding | — | 72.2% |
| HumanEval+ | 82.3% | — |
| MBPP+ | 75.1% | — |
Reasoning Too close to call
Deepseek Coder v2: 23.6 (#176), QwQ-32B: 23.7 (#174)
| Benchmark | Deepseek Coder v2 | QwQ-32B |
|---|---|---|
| LMArena Hard Prompts | 1207 | 1325 |
| Chess Puzzles | — | 5% |
| LiveBench Reasoning | — | 83.5% |
| LiveBench Data Analysis | — | 65% |
| Epoch Capabilities Index | — | 137.6 |
| ForecastBench | — | 58.3 |
| LiveBench | — | 72% |
| WinoGrande | 83.7% | — |
Math QwQ-32B leads
Deepseek Coder v2: 34.9 (#190), QwQ-32B: 38.0 (#143)
| Benchmark | Deepseek Coder v2 | QwQ-32B |
|---|---|---|
| LMArena Math | 1241 | 1359 |
| OTIS Mock AIME 2024-2025 | — | 59.2% |
| LiveBench Math | — | 77.8% |
| GSM8K | 94.5% | — |
Knowledge QwQ-32B leads
Deepseek Coder v2: 32.3 (#212), QwQ-32B: 37.2 (#158)
| Benchmark | Deepseek Coder v2 | QwQ-32B |
|---|---|---|
| LMArena Expert | 1181 | 1324 |
| GPQA Diamond | — | 65.3% |
| Confabulations | — | 15.6% |
| ARC (AI2) Challenge | 64.3% | — |
Multilingual QwQ-32B leads
Deepseek Coder v2: 36.3 (#240), QwQ-32B: 44.8 (#176)
| Benchmark | Deepseek Coder v2 | QwQ-32B |
|---|---|---|
| LMArena Non-English | 1182 | 1305 |
| LMArena Chinese | 1201 | 1378 |
| LMArena French | 1185 | 1336 |
| LMArena German | 1164 | 1313 |
| LMArena Japanese | 1126 | 1262 |
| LMArena Korean | 1104 | 1279 |
| LMArena Russian | 1188 | 1297 |
| LMArena Spanish | 1153 | 1354 |
Instruction Following QwQ-32B leads
Deepseek Coder v2: 61.7 (#242), QwQ-32B: 72.6 (#137)
| Benchmark | Deepseek Coder v2 | QwQ-32B |
|---|---|---|
| LMArena Instruction Following | 1180 | 1297 |
| LiveBench Instruction Following | — | 81.8% |
Long Context QwQ-32B leads
Deepseek Coder v2: 37.0 (#224), QwQ-32B: 49.0 (#11)
| Benchmark | Deepseek Coder v2 | QwQ-32B |
|---|---|---|
| LMArena Longer Query | 1219 | 1308 |
| Fiction.LiveBench | — | 83.3% |
Writing & Preference QwQ-32B leads
Deepseek Coder v2: 38.2 (#253), QwQ-32B: 50.6 (#180)
| Benchmark | Deepseek Coder v2 | QwQ-32B |
|---|---|---|
| LMArena Text | 1191 | 1329 |
| LMArena Creative Writing | 1120 | 1288 |
| LMArena Multi-Turn | 1177 | 1314 |
| Short-Story Creative Writing | — | 80.2% |
| EQ-Bench Creative Writing | — | 1257 |
| LiveBench Language | — | 51.4% |
Frequently asked questions
Is Deepseek Coder v2 better than QwQ-32B?
QwQ-32B is the stronger model overall, scoring 39.8 to 35.9 on the Noometry Index.
Is Deepseek Coder v2 or QwQ-32B better for coding?
Deepseek Coder v2 scores higher on coding benchmarks: 38.1 versus 35.4 in the Noometry coding category.
How many benchmarks do Deepseek Coder v2 and QwQ-32B share?
19 benchmarks have published results for both models. Deepseek Coder v2 has 24 scored results on Noometry and QwQ-32B has 36.