Model comparison
DeepSeek LLM 67B vs Qwen-14B
Qwen-14B is the stronger model overall, scoring 30.8 to 24.8 on the Noometry Index.
Last verified . 11 shared benchmarks.
DeepSeek
Alibaba (Qwen)Summary
- They share 11 benchmarks with published results for both. DeepSeek LLM 67B scores higher in 5 categories and Qwen-14B in 2 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen-14B leads 28.7 to 8.1.
Side by side
| DeepSeek LLM 67B | Qwen-14B | |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 24.8 | 30.8 |
| Released | 2023-11-29 | 2023-09-24 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 15 | 18 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
DeepSeek LLM 67B: 31.3 (#273), Qwen-14B: 30.6 (#283)
| Benchmark | DeepSeek LLM 67B | Qwen-14B |
|---|---|---|
| LMArena Coding | 1096 | 1072 |
Reasoning Qwen-14B leads
DeepSeek LLM 67B: 15.9 (#300), Qwen-14B: 18.7 (#255)
| Benchmark | DeepSeek LLM 67B | Qwen-14B |
|---|---|---|
| LMArena Hard Prompts | 1070 | 1027 |
| Epoch Capabilities Index | 110.54 | 112.8 |
| Chess Puzzles | 0% | — |
| BIG-Bench Hard | — | 55% |
| LAMBADA | — | 71.1% |
| PIQA | — | 79.9% |
Math Qwen-14B leads
DeepSeek LLM 67B: 8.1 (#318), Qwen-14B: 28.7 (#224)
| Benchmark | DeepSeek LLM 67B | Qwen-14B |
|---|---|---|
| LMArena Math | 1108 | 1068 |
| OTIS Mock AIME 2024-2025 | 0.8% | — |
| MATH Level 5 | 6.4% | — |
| GSM8K | — | 61.3% |
Knowledge Not comparable
DeepSeek LLM 67B: 7.1 (#308), Qwen-14B: —
| Benchmark | DeepSeek LLM 67B | Qwen-14B |
|---|---|---|
| GPQA Diamond | 24.6% | — |
| ARC (AI2) Challenge | — | 84.4% |
| BoolQ | — | 86.2% |
| MMLU | — | 66.3% |
Multilingual DeepSeek LLM 67B leads
DeepSeek LLM 67B: 29.5 (#262), Qwen-14B: 27.7 (#270)
| Benchmark | DeepSeek LLM 67B | Qwen-14B |
|---|---|---|
| LMArena Non-English | 1073 | 1042 |
| LMArena Chinese | 1132 | 1077 |
Instruction Following DeepSeek LLM 67B leads
DeepSeek LLM 67B: 55.7 (#272), Qwen-14B: 52.6 (#284)
| Benchmark | DeepSeek LLM 67B | Qwen-14B |
|---|---|---|
| LMArena Instruction Following | 1079 | 1031 |
Long Context DeepSeek LLM 67B leads
DeepSeek LLM 67B: 33.1 (#260), Qwen-14B: 31.3 (#276)
| Benchmark | DeepSeek LLM 67B | Qwen-14B |
|---|---|---|
| LMArena Longer Query | 1092 | 1029 |
Writing & Preference DeepSeek LLM 67B leads
DeepSeek LLM 67B: 31.8 (#277), Qwen-14B: 27.8 (#294)
| Benchmark | DeepSeek LLM 67B | Qwen-14B |
|---|---|---|
| LMArena Text | 1105 | 1051 |
| LMArena Creative Writing | 1067 | 1028 |
| LMArena Multi-Turn | 1082 | 1022 |
Frequently asked questions
Is DeepSeek LLM 67B better than Qwen-14B?
Qwen-14B is the stronger model overall, scoring 30.8 to 24.8 on the Noometry Index.
Is DeepSeek LLM 67B or Qwen-14B better for coding?
They score almost the same on coding (31.3 vs 30.6); test both on your own repository before choosing.
How many benchmarks do DeepSeek LLM 67B and Qwen-14B share?
11 benchmarks have published results for both models. DeepSeek LLM 67B has 15 scored results on Noometry and Qwen-14B has 18.