Model comparison
Deepseek Coder v2 vs Qwen-14B
Deepseek Coder v2 is the stronger model overall, scoring 35.9 to 31.4 on the Noometry Index.
Last verified . 12 shared benchmarks.
DeepSeek
Alibaba (Qwen)Summary
- They share 12 benchmarks with published results for both. Deepseek Coder v2 scores higher in 7 categories and Qwen-14B in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Deepseek Coder v2 leads 38.2 to 27.6.
Side by side
| Deepseek Coder v2 | Qwen-14B | |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 35.9 | 31.4 |
| Released | 2024-06-17 | 2023-09-24 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 24 | 18 |
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), Qwen-14B: 31.2 (#288)
| Benchmark | Deepseek Coder v2 | Qwen-14B |
|---|---|---|
| LMArena Coding | 1251 | 1071 |
| BigCodeBench Instruct | 48.2% | — |
| BigCodeBench Complete | 59.7% | — |
| HumanEval+ | 82.3% | — |
| MBPP+ | 75.1% | — |
Reasoning Deepseek Coder v2 leads
Deepseek Coder v2: 23.6 (#176), Qwen-14B: 19.6 (#257)
| Benchmark | Deepseek Coder v2 | Qwen-14B |
|---|---|---|
| LMArena Hard Prompts | 1207 | 1027 |
| BIG-Bench Hard | — | 55% |
| Epoch Capabilities Index | — | 113.03 |
| LAMBADA | — | 71.1% |
| PIQA | — | 79.9% |
| WinoGrande | 83.7% | — |
Math Deepseek Coder v2 leads
Deepseek Coder v2: 34.9 (#190), Qwen-14B: 31.2 (#227)
| Benchmark | Deepseek Coder v2 | Qwen-14B |
|---|---|---|
| LMArena Math | 1241 | 1068 |
| GSM8K | 94.5% | 61.3% |
Knowledge Not comparable
Deepseek Coder v2: 32.3 (#212), Qwen-14B: —
| Benchmark | Deepseek Coder v2 | Qwen-14B |
|---|---|---|
| ARC (AI2) Challenge | 64.3% | 84.4% |
| LMArena Expert | 1181 | — |
| BoolQ | — | 86.2% |
| MMLU | — | 66.3% |
Multilingual Deepseek Coder v2 leads
Deepseek Coder v2: 36.3 (#240), Qwen-14B: 27.5 (#275)
| Benchmark | Deepseek Coder v2 | Qwen-14B |
|---|---|---|
| LMArena Non-English | 1182 | 1041 |
| LMArena Chinese | 1201 | 1077 |
| LMArena French | 1185 | — |
| LMArena German | 1164 | — |
| LMArena Japanese | 1126 | — |
| LMArena Korean | 1104 | — |
| LMArena Russian | 1188 | — |
| LMArena Spanish | 1153 | — |
Instruction Following Deepseek Coder v2 leads
Deepseek Coder v2: 61.7 (#242), Qwen-14B: 52.4 (#289)
| Benchmark | Deepseek Coder v2 | Qwen-14B |
|---|---|---|
| LMArena Instruction Following | 1180 | 1031 |
Long Context Deepseek Coder v2 leads
Deepseek Coder v2: 37.0 (#224), Qwen-14B: 31.3 (#280)
| Benchmark | Deepseek Coder v2 | Qwen-14B |
|---|---|---|
| LMArena Longer Query | 1219 | 1028 |
Writing & Preference Deepseek Coder v2 leads
Deepseek Coder v2: 38.2 (#253), Qwen-14B: 27.6 (#299)
| Benchmark | Deepseek Coder v2 | Qwen-14B |
|---|---|---|
| LMArena Text | 1191 | 1051 |
| LMArena Creative Writing | 1120 | 1028 |
| LMArena Multi-Turn | 1177 | 1022 |
Frequently asked questions
Is Deepseek Coder v2 better than Qwen-14B?
Deepseek Coder v2 is the stronger model overall, scoring 35.9 to 31.4 on the Noometry Index.
Is Deepseek Coder v2 or Qwen-14B better for coding?
Deepseek Coder v2 scores higher on coding benchmarks: 38.1 versus 31.2 in the Noometry coding category.
How many benchmarks do Deepseek Coder v2 and Qwen-14B share?
12 benchmarks have published results for both models. Deepseek Coder v2 has 24 scored results on Noometry and Qwen-14B has 18.