Model comparison
DeepSeek-V2.5 (Sep 2024) vs Qwen1.5-32B
DeepSeek-V2.5 (Sep 2024) is the stronger model overall, scoring 37.6 to 30.5 on the Noometry Index.
Last verified . 19 shared benchmarks.
DeepSeek
Alibaba (Qwen)Summary
- They share 19 benchmarks with published results for both. DeepSeek-V2.5 (Sep 2024) scores higher in 8 categories and Qwen1.5-32B in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where DeepSeek-V2.5 (Sep 2024) leads 34.8 to 13.5.
- The biggest single-benchmark swing is BigCodeBench Instruct: 48.6% for DeepSeek-V2.5 (Sep 2024) and 32.3% for Qwen1.5-32B.
Side by side
| DeepSeek-V2.5 (Sep 2024) | Qwen1.5-32B | |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 37.6 | 30.5 |
| Released | 2024-09-06 | 2024-02-04 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 22 | 21 |
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), Qwen1.5-32B: 31.7 (#282)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen1.5-32B |
|---|---|---|
| BigCodeBench Instruct | 48.6% | 32.3% |
| LMArena Coding | 1309 | 1155 |
| BigCodeBench Complete | 53.2% | 42% |
| Aider Polyglot | 17.8% | — |
| HumanEval+ | 83.5% | — |
| MBPP+ | 74.1% | — |
Reasoning DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 25.6 (#145), Qwen1.5-32B: 21.8 (#212)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen1.5-32B |
|---|---|---|
| LMArena Hard Prompts | 1289 | 1130 |
Math DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 35.9 (#177), Qwen1.5-32B: 33.0 (#207)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen1.5-32B |
|---|---|---|
| LMArena Math | 1288 | 1155 |
Knowledge DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 34.8 (#193), Qwen1.5-32B: 13.5 (#296)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen1.5-32B |
|---|---|---|
| LMArena Expert | 1266 | 1126 |
| GPQA Diamond | — | 30.7% |
| MMLU | — | 74.4% |
Multilingual DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 42.5 (#193), Qwen1.5-32B: 31.4 (#259)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen1.5-32B |
|---|---|---|
| LMArena Non-English | 1273 | 1106 |
| LMArena Chinese | 1318 | 1177 |
| LMArena French | 1289 | 1101 |
| LMArena German | 1258 | 1058 |
| LMArena Japanese | 1228 | 1027 |
| LMArena Korean | 1209 | 1008 |
| LMArena Russian | 1289 | 1073 |
| LMArena Spanish | 1248 | 1089 |
Instruction Following DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 67.5 (#194), Qwen1.5-32B: 57.7 (#265)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen1.5-32B |
|---|---|---|
| LMArena Instruction Following | 1280 | 1116 |
Long Context DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 39.5 (#174), Qwen1.5-32B: 34.7 (#246)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen1.5-32B |
|---|---|---|
| LMArena Longer Query | 1301 | 1146 |
Writing & Preference DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 49.8 (#187), Qwen1.5-32B: 34.2 (#271)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen1.5-32B |
|---|---|---|
| LMArena Text | 1294 | 1137 |
| LMArena Creative Writing | 1285 | 1083 |
| LMArena Multi-Turn | 1297 | 1140 |
Frequently asked questions
Is DeepSeek-V2.5 (Sep 2024) better than Qwen1.5-32B?
DeepSeek-V2.5 (Sep 2024) is the stronger model overall, scoring 37.6 to 30.5 on the Noometry Index.
Is DeepSeek-V2.5 (Sep 2024) or Qwen1.5-32B better for coding?
They score almost the same on coding (31.7 vs 31.7); test both on your own repository before choosing.
How many benchmarks do DeepSeek-V2.5 (Sep 2024) and Qwen1.5-32B share?
19 benchmarks have published results for both models. DeepSeek-V2.5 (Sep 2024) has 22 scored results on Noometry and Qwen1.5-32B has 21.