# Deepseek Coder v2 vs Qwen3.5 27B

> Qwen3.5 27B is the stronger model overall, scoring 41.9 to 35.9 on the Noometry Index.

- Canonical page: https://noometry.com/compare/deepseek-coder-v2-vs-qwen3-5-27b
- Last updated: 2026-10-10
- Shared benchmarks: 17

## Summary

- They share 17 benchmarks with published results for both. Deepseek Coder v2 scores higher in 0 categories and Qwen3.5 27B in 8 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Qwen3.5 27B leads 59.3 to 38.2.

## Snapshot

| | Deepseek Coder v2 | Qwen3.5 27B |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 35.9 | 41.9 |
| Rank | 220 | 127 |
| Context | — | 262K |
| Input $/M | — | $0.30 |
| Output $/M | — | $2.40 |
| Weights | Open | Open |

## Coding

- Deepseek Coder v2: 38.1 (#183)
- Qwen3.5 27B: 38.9 (#168)

| Benchmark | Deepseek Coder v2 | Qwen3.5 27B |
|---|---|---|
| LMArena Coding | 1251 | 1427 |
| LMArena WebDev | — | 1358 |
| WeirdML | — | 39.5% |
| BigCodeBench Instruct | 48.2% | — |
| BigCodeBench Complete | 59.7% | — |
| ALE-Bench | — | 349.45 |
| HumanEval+ | 82.3% | — |
| MBPP+ | 75.1% | — |

## Agentic & Tool Use

- Deepseek Coder v2: —
- Qwen3.5 27B: —

| Benchmark | Deepseek Coder v2 | Qwen3.5 27B |
|---|---|---|
| Vending-Bench 2 | — | 201.98 |

## Reasoning

- Deepseek Coder v2: 23.6 (#176)
- Qwen3.5 27B: 27.5 (#117)

| Benchmark | Deepseek Coder v2 | Qwen3.5 27B |
|---|---|---|
| LMArena Hard Prompts | 1207 | 1414 |
| NYT Connections (extended) | — | 47.9% |
| Thematic Generalization | — | 45.5% |
| DTBench | — | 82.4% |
| LMCA | — | 34% |
| WinoGrande | 83.7% | — |

## Math

- Deepseek Coder v2: 34.9 (#190)
- Qwen3.5 27B: 38.8 (#127)

| Benchmark | Deepseek Coder v2 | Qwen3.5 27B |
|---|---|---|
| LMArena Math | 1241 | 1429 |
| MathArena Final-Answer Competitions | — | 56.7% |
| GSM8K | 94.5% | — |

## Knowledge

- Deepseek Coder v2: 32.3 (#212)
- Qwen3.5 27B: 38.0 (#150)

| Benchmark | Deepseek Coder v2 | Qwen3.5 27B |
|---|---|---|
| LMArena Expert | 1181 | 1428 |
| Vectara Hallucination Rate | — | 12.1% |
| ARC (AI2) Challenge | 64.3% | — |

## Multimodal

- Deepseek Coder v2: —
- Qwen3.5 27B: 39.4 (#59)

| Benchmark | Deepseek Coder v2 | Qwen3.5 27B |
|---|---|---|
| LMArena Vision | — | 1241 |

## Multilingual

- Deepseek Coder v2: 36.3 (#240)
- Qwen3.5 27B: 50.8 (#115)

| Benchmark | Deepseek Coder v2 | Qwen3.5 27B |
|---|---|---|
| LMArena Non-English | 1182 | 1390 |
| LMArena Chinese | 1201 | 1478 |
| LMArena French | 1185 | 1410 |
| LMArena German | 1164 | 1393 |
| LMArena Japanese | 1126 | 1345 |
| LMArena Korean | 1104 | 1358 |
| LMArena Russian | 1188 | 1390 |
| LMArena Spanish | 1153 | 1407 |

## Instruction Following

- Deepseek Coder v2: 61.7 (#242)
- Qwen3.5 27B: 73.5 (#119)

| Benchmark | Deepseek Coder v2 | Qwen3.5 27B |
|---|---|---|
| LMArena Instruction Following | 1180 | 1393 |

## Long Context

- Deepseek Coder v2: 37.0 (#224)
- Qwen3.5 27B: 43.1 (#106)

| Benchmark | Deepseek Coder v2 | Qwen3.5 27B |
|---|---|---|
| LMArena Longer Query | 1219 | 1413 |

## Writing & Preference

- Deepseek Coder v2: 38.2 (#253)
- Qwen3.5 27B: 59.3 (#111)

| Benchmark | Deepseek Coder v2 | Qwen3.5 27B |
|---|---|---|
| LMArena Text | 1191 | 1409 |
| LMArena Creative Writing | 1120 | 1362 |
| LMArena Multi-Turn | 1177 | 1410 |

## FAQ

### Is Deepseek Coder v2 better than Qwen3.5 27B?

Qwen3.5 27B is the stronger model overall, scoring 41.9 to 35.9 on the Noometry Index.

### Is Deepseek Coder v2 or Qwen3.5 27B better for coding?

They score almost the same on coding (38.1 vs 38.9); test both on your own repository before choosing.

### How many benchmarks do Deepseek Coder v2 and Qwen3.5 27B share?

17 benchmarks have published results for both models. Deepseek Coder v2 has 24 scored results on Noometry and Qwen3.5 27B has 28.