Model comparison
DeepSeek-V2 (MoE-236B, May 2024) vs Qwen-14B
Qwen-14B has enough public results to be ranked (#275); DeepSeek-V2 (MoE-236B, May 2024) does not yet, so treat this comparison as directional.
Last verified . 5 shared benchmarks.
DeepSeek
Alibaba (Qwen)Summary
- They share 5 benchmarks with published results for both. DeepSeek-V2 (MoE-236B, May 2024) scores higher in 1 category and Qwen-14B in 0 categories; one gap is clear of the uncertainty.
- The widest gap is in coding, where DeepSeek-V2 (MoE-236B, May 2024) leads 40.4 to 31.2.
Side by side
| DeepSeek-V2 (MoE-236B, May 2024) | Qwen-14B | |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 40.3 | 31.4 |
| Released | 2024-05-07 | 2023-09-24 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 10 | 18 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-V2 (MoE-236B, May 2024) leads
DeepSeek-V2 (MoE-236B, May 2024): 40.4 (#139), Qwen-14B: 31.2 (#288)
| Benchmark | DeepSeek-V2 (MoE-236B, May 2024) | Qwen-14B |
|---|---|---|
| BigCodeBench Instruct | 48.9% | — |
| LMArena Coding | — | 1071 |
| BigCodeBench Complete | 59.4% | — |
Reasoning Not comparable
DeepSeek-V2 (MoE-236B, May 2024): —, Qwen-14B: 19.6 (#257)
| Benchmark | DeepSeek-V2 (MoE-236B, May 2024) | Qwen-14B |
|---|---|---|
| BIG-Bench Hard | 78.8% | 55% |
| Epoch Capabilities Index | 124.77 | 113.03 |
| PIQA | 83.9% | 79.9% |
| LMArena Hard Prompts | — | 1027 |
| HellaSwag | 87.1% | — |
| LAMBADA | — | 71.1% |
| WinoGrande | 86.3% | — |
Math Not comparable
DeepSeek-V2 (MoE-236B, May 2024): —, Qwen-14B: 31.2 (#227)
| Benchmark | DeepSeek-V2 (MoE-236B, May 2024) | Qwen-14B |
|---|---|---|
| LMArena Math | — | 1068 |
| GSM8K | — | 61.3% |
Knowledge Not comparable
DeepSeek-V2 (MoE-236B, May 2024): —, Qwen-14B: —
| Benchmark | DeepSeek-V2 (MoE-236B, May 2024) | Qwen-14B |
|---|---|---|
| ARC (AI2) Challenge | 92.2% | 84.4% |
| MMLU | 78.4% | 66.3% |
| BoolQ | — | 86.2% |
| TriviaQA | 80% | — |
Multilingual Not comparable
DeepSeek-V2 (MoE-236B, May 2024): —, Qwen-14B: 27.5 (#275)
| Benchmark | DeepSeek-V2 (MoE-236B, May 2024) | Qwen-14B |
|---|---|---|
| LMArena Non-English | — | 1041 |
| LMArena Chinese | — | 1077 |
Instruction Following Not comparable
DeepSeek-V2 (MoE-236B, May 2024): —, Qwen-14B: 52.4 (#289)
| Benchmark | DeepSeek-V2 (MoE-236B, May 2024) | Qwen-14B |
|---|---|---|
| LMArena Instruction Following | — | 1031 |
Long Context Not comparable
DeepSeek-V2 (MoE-236B, May 2024): —, Qwen-14B: 31.3 (#280)
| Benchmark | DeepSeek-V2 (MoE-236B, May 2024) | Qwen-14B |
|---|---|---|
| LMArena Longer Query | — | 1028 |
Writing & Preference Not comparable
DeepSeek-V2 (MoE-236B, May 2024): —, Qwen-14B: 27.6 (#299)
| Benchmark | DeepSeek-V2 (MoE-236B, May 2024) | Qwen-14B |
|---|---|---|
| LMArena Text | — | 1051 |
| LMArena Creative Writing | — | 1028 |
| LMArena Multi-Turn | — | 1022 |
Frequently asked questions
Is DeepSeek-V2 (MoE-236B, May 2024) better than Qwen-14B?
Qwen-14B has enough public results to be ranked (#275); DeepSeek-V2 (MoE-236B, May 2024) does not yet, so treat this comparison as directional.
Is DeepSeek-V2 (MoE-236B, May 2024) or Qwen-14B better for coding?
DeepSeek-V2 (MoE-236B, May 2024) scores higher on coding benchmarks: 40.4 versus 31.2 in the Noometry coding category.
How many benchmarks do DeepSeek-V2 (MoE-236B, May 2024) and Qwen-14B share?
5 benchmarks have published results for both models. DeepSeek-V2 (MoE-236B, May 2024) has 10 scored results on Noometry and Qwen-14B has 18.