# Deepseek Coder v2 vs Qwen3.5-Flash

> Qwen3.5-Flash is the stronger model overall, scoring 42.5 to 35.9 on the Noometry Index.

- Canonical page: https://noometry.com/compare/deepseek-coder-v2-vs-qwen3-5-flash
- Last updated: 2026-10-11
- Shared benchmarks: 17

## Summary

- They share 17 benchmarks with published results for both. Deepseek Coder v2 scores higher in 1 category and Qwen3.5-Flash in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Qwen3.5-Flash leads 57.9 to 38.2.
- Deepseek Coder v2 has downloadable open weights; the other is API-only.

## Snapshot

| | Deepseek Coder v2 | Qwen3.5-Flash |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 35.9 | 42.5 |
| Rank | 220 | 112 |
| Context | — | 1M |
| Input $/M | — | $0.10 |
| Output $/M | — | $0.40 |
| Weights | Open | Proprietary |

## Coding

- Deepseek Coder v2: 38.1 (#183)
- Qwen3.5-Flash: 34.2 (#242)

| Benchmark | Deepseek Coder v2 | Qwen3.5-Flash |
|---|---|---|
| LMArena Coding | 1251 | 1412 |
| LMArena WebDev | — | 1244 |
| BigCodeBench Instruct | 48.2% | — |
| BigCodeBench Complete | 59.7% | — |
| ALE-Bench | — | 221.8 |
| HumanEval+ | 82.3% | — |
| MBPP+ | 75.1% | — |

## Agentic & Tool Use

- Deepseek Coder v2: —
- Qwen3.5-Flash: —

| Benchmark | Deepseek Coder v2 | Qwen3.5-Flash |
|---|---|---|
| Vending-Bench 2 | — | 462.69 |

## Reasoning

- Deepseek Coder v2: 23.6 (#176)
- Qwen3.5-Flash: 33.7 (#72)

| Benchmark | Deepseek Coder v2 | Qwen3.5-Flash |
|---|---|---|
| LMArena Hard Prompts | 1207 | 1403 |
| Chess Puzzles | — | 21% |
| Mystery Game Puzzles | — | 20% |
| DTBench | — | 82.9% |
| LMCA | — | 29.1% |
| Epoch Capabilities Index | — | 143.98 |
| WinoGrande | 83.7% | — |

## Math

- Deepseek Coder v2: 34.9 (#190)
- Qwen3.5-Flash: 37.4 (#158)

| Benchmark | Deepseek Coder v2 | Qwen3.5-Flash |
|---|---|---|
| LMArena Math | 1241 | 1407 |
| FrontierMath (Tiers 1-3) | — | 18.2% |
| OTIS Mock AIME 2024-2025 | — | 84.4% |
| FrontierMath (Feb 2025 set) | — | 6.2% |
| FrontierMath Tier 4 (v1) | — | 0% |
| GSM8K | 94.5% | — |

## Knowledge

- Deepseek Coder v2: 32.3 (#212)
- Qwen3.5-Flash: 43.2 (#93)

| Benchmark | Deepseek Coder v2 | Qwen3.5-Flash |
|---|---|---|
| LMArena Expert | 1181 | 1407 |
| GPQA Diamond | — | 82.3% |
| SimpleQA Verified | — | 20.3% |
| Vectara Hallucination Rate | — | 10.5% |
| ARC (AI2) Challenge | 64.3% | — |

## Multilingual

- Deepseek Coder v2: 36.3 (#240)
- Qwen3.5-Flash: 50.5 (#121)

| Benchmark | Deepseek Coder v2 | Qwen3.5-Flash |
|---|---|---|
| LMArena Non-English | 1182 | 1385 |
| LMArena Chinese | 1201 | 1446 |
| LMArena French | 1185 | 1412 |
| LMArena German | 1164 | 1390 |
| LMArena Japanese | 1126 | 1368 |
| LMArena Korean | 1104 | 1344 |
| LMArena Russian | 1188 | 1379 |
| LMArena Spanish | 1153 | 1400 |

## Instruction Following

- Deepseek Coder v2: 61.7 (#242)
- Qwen3.5-Flash: 72.6 (#139)

| Benchmark | Deepseek Coder v2 | Qwen3.5-Flash |
|---|---|---|
| LMArena Instruction Following | 1180 | 1374 |

## Long Context

- Deepseek Coder v2: 37.0 (#224)
- Qwen3.5-Flash: 42.4 (#124)

| Benchmark | Deepseek Coder v2 | Qwen3.5-Flash |
|---|---|---|
| LMArena Longer Query | 1219 | 1392 |

## Writing & Preference

- Deepseek Coder v2: 38.2 (#253)
- Qwen3.5-Flash: 57.9 (#122)

| Benchmark | Deepseek Coder v2 | Qwen3.5-Flash |
|---|---|---|
| LMArena Text | 1191 | 1397 |
| LMArena Creative Writing | 1120 | 1343 |
| LMArena Multi-Turn | 1177 | 1393 |

## FAQ

### Is Deepseek Coder v2 better than Qwen3.5-Flash?

Qwen3.5-Flash is the stronger model overall, scoring 42.5 to 35.9 on the Noometry Index.

### Is Deepseek Coder v2 or Qwen3.5-Flash better for coding?

Deepseek Coder v2 scores higher on coding benchmarks: 38.1 versus 34.2 in the Noometry coding category.

### How many benchmarks do Deepseek Coder v2 and Qwen3.5-Flash share?

17 benchmarks have published results for both models. Deepseek Coder v2 has 24 scored results on Noometry and Qwen3.5-Flash has 32.