# DeepSeek-V2.5 (Sep 2024) vs Qwen1.5-72B

> DeepSeek-V2.5 (Sep 2024) is the stronger model overall, scoring 37.0 to 30.3 on the Noometry Index.

- Canonical page: https://noometry.com/compare/deepseek-v2-5-vs-qwen1-5-72b
- Last updated: 2026-10-10
- Shared benchmarks: 21

## Summary

- They share 21 benchmarks with published results for both. DeepSeek-V2.5 (Sep 2024) scores higher in 7 categories and Qwen1.5-72B in 1 category; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where DeepSeek-V2.5 (Sep 2024) leads 34.2 to 11.5.
- The biggest single-benchmark swing is BigCodeBench Instruct: 48.6% for DeepSeek-V2.5 (Sep 2024) and 33.2% for Qwen1.5-72B.

## Snapshot

| | DeepSeek-V2.5 (Sep 2024) | Qwen1.5-72B |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 37.0 | 30.3 |
| Rank | 197 | 282 |
| Context | — | — |
| Input $/M | — | — |
| Output $/M | — | — |
| Weights | Open | Open |

## Coding

- DeepSeek-V2.5 (Sep 2024): 31.3 (#274)
- Qwen1.5-72B: 31.4 (#272)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen1.5-72B |
|---|---|---|
| BigCodeBench Instruct | 48.6% | 33.2% |
| LMArena Coding | 1310 | 1166 |
| BigCodeBench Complete | 53.2% | 40.3% |
| HumanEval+ | 83.5% | 59.1% |
| MBPP+ | 74.1% | 61.6% |
| Aider Polyglot | 17.8% | — |

## Reasoning

- DeepSeek-V2.5 (Sep 2024): 24.5 (#145)
- Qwen1.5-72B: 21.2 (#201)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen1.5-72B |
|---|---|---|
| LMArena Hard Prompts | 1289 | 1148 |

## Math

- DeepSeek-V2.5 (Sep 2024): 33.3 (#177)
- Qwen1.5-72B: 30.7 (#205)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen1.5-72B |
|---|---|---|
| LMArena Math | 1289 | 1164 |

## Knowledge

- DeepSeek-V2.5 (Sep 2024): 34.2 (#190)
- Qwen1.5-72B: 11.5 (#295)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen1.5-72B |
|---|---|---|
| LMArena Expert | 1266 | 1136 |
| GPQA Diamond | — | 28.8% |

## Multilingual

- DeepSeek-V2.5 (Sep 2024): 42.7 (#188)
- Qwen1.5-72B: 33.4 (#248)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen1.5-72B |
|---|---|---|
| LMArena Non-English | 1273 | 1135 |
| LMArena Chinese | 1317 | 1186 |
| LMArena French | 1288 | 1158 |
| LMArena German | 1258 | 1084 |
| LMArena Japanese | 1228 | 1061 |
| LMArena Korean | 1209 | 1049 |
| LMArena Russian | 1289 | 1104 |
| LMArena Spanish | 1248 | 1110 |

## Instruction Following

- DeepSeek-V2.5 (Sep 2024): 67.7 (#189)
- Qwen1.5-72B: 59.6 (#251)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen1.5-72B |
|---|---|---|
| LMArena Instruction Following | 1280 | 1142 |

## Long Context

- DeepSeek-V2.5 (Sep 2024): 39.5 (#169)
- Qwen1.5-72B: 35.0 (#238)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen1.5-72B |
|---|---|---|
| LMArena Longer Query | 1301 | 1157 |

## Writing & Preference

- DeepSeek-V2.5 (Sep 2024): 50.0 (#182)
- Qwen1.5-72B: 37.4 (#253)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen1.5-72B |
|---|---|---|
| LMArena Text | 1294 | 1166 |
| LMArena Creative Writing | 1285 | 1137 |
| LMArena Multi-Turn | 1297 | 1160 |

## FAQ

### Is DeepSeek-V2.5 (Sep 2024) better than Qwen1.5-72B?

DeepSeek-V2.5 (Sep 2024) is the stronger model overall, scoring 37.0 to 30.3 on the Noometry Index.

### Is DeepSeek-V2.5 (Sep 2024) or Qwen1.5-72B better for coding?

They score almost the same on coding (31.3 vs 31.4); test both on your own repository before choosing.

### How many benchmarks do DeepSeek-V2.5 (Sep 2024) and Qwen1.5-72B share?

21 benchmarks have published results for both models. DeepSeek-V2.5 (Sep 2024) has 22 scored results on Noometry and Qwen1.5-72B has 22.