Model comparison
DeepSeek-V2.5 (Sep 2024) vs Qwen-14B
DeepSeek-V2.5 (Sep 2024) is the stronger model overall, scoring 37.6 to 31.4 on the Noometry Index.
Last verified . 10 shared benchmarks.
DeepSeek
Alibaba (Qwen)Summary
- They share 10 benchmarks with published results for both. DeepSeek-V2.5 (Sep 2024) scores higher in 7 categories and Qwen-14B in 0 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where DeepSeek-V2.5 (Sep 2024) leads 49.8 to 27.6.
Side by side
| DeepSeek-V2.5 (Sep 2024) | Qwen-14B | |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 37.6 | 31.4 |
| Released | 2024-09-06 | 2023-09-24 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 22 | 18 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
DeepSeek-V2.5 (Sep 2024): 31.7 (#281), Qwen-14B: 31.2 (#288)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen-14B |
|---|---|---|
| LMArena Coding | 1309 | 1071 |
| Aider Polyglot | 17.8% | — |
| BigCodeBench Instruct | 48.6% | — |
| BigCodeBench Complete | 53.2% | — |
| HumanEval+ | 83.5% | — |
| MBPP+ | 74.1% | — |
Reasoning DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 25.6 (#145), Qwen-14B: 19.6 (#257)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen-14B |
|---|---|---|
| LMArena Hard Prompts | 1289 | 1027 |
| BIG-Bench Hard | — | 55% |
| Epoch Capabilities Index | — | 113.03 |
| LAMBADA | — | 71.1% |
| PIQA | — | 79.9% |
Math DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 35.9 (#177), Qwen-14B: 31.2 (#227)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen-14B |
|---|---|---|
| LMArena Math | 1288 | 1068 |
| GSM8K | — | 61.3% |
Knowledge Not comparable
DeepSeek-V2.5 (Sep 2024): 34.8 (#193), Qwen-14B: —
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen-14B |
|---|---|---|
| LMArena Expert | 1266 | — |
| ARC (AI2) Challenge | — | 84.4% |
| BoolQ | — | 86.2% |
| MMLU | — | 66.3% |
Multilingual DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 42.5 (#193), Qwen-14B: 27.5 (#275)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen-14B |
|---|---|---|
| LMArena Non-English | 1273 | 1041 |
| LMArena Chinese | 1318 | 1077 |
| LMArena French | 1289 | — |
| LMArena German | 1258 | — |
| LMArena Japanese | 1228 | — |
| LMArena Korean | 1209 | — |
| LMArena Russian | 1289 | — |
| LMArena Spanish | 1248 | — |
Instruction Following DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 67.5 (#194), Qwen-14B: 52.4 (#289)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen-14B |
|---|---|---|
| LMArena Instruction Following | 1280 | 1031 |
Long Context DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 39.5 (#174), Qwen-14B: 31.3 (#280)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen-14B |
|---|---|---|
| LMArena Longer Query | 1301 | 1028 |
Writing & Preference DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 49.8 (#187), Qwen-14B: 27.6 (#299)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen-14B |
|---|---|---|
| LMArena Text | 1294 | 1051 |
| LMArena Creative Writing | 1285 | 1028 |
| LMArena Multi-Turn | 1297 | 1022 |
Frequently asked questions
Is DeepSeek-V2.5 (Sep 2024) better than Qwen-14B?
DeepSeek-V2.5 (Sep 2024) is the stronger model overall, scoring 37.6 to 31.4 on the Noometry Index.
Is DeepSeek-V2.5 (Sep 2024) or Qwen-14B better for coding?
They score almost the same on coding (31.7 vs 31.2); test both on your own repository before choosing.
How many benchmarks do DeepSeek-V2.5 (Sep 2024) and Qwen-14B share?
10 benchmarks have published results for both models. DeepSeek-V2.5 (Sep 2024) has 22 scored results on Noometry and Qwen-14B has 18.