# DeepSeek-V2 (MoE-236B, May 2024) vs DeepSeek-V3.2-Speciale

> DeepSeek-V3.2-Speciale has enough public results to be ranked (#159); DeepSeek-V2 (MoE-236B, May 2024) does not yet, so treat this comparison as directional.

- Canonical page: https://noometry.com/compare/deepseek-v2-vs-deepseek-v3-2-speciale
- Last updated: 2026-10-10
- Shared benchmarks: 0

## Snapshot

| | DeepSeek-V2 (MoE-236B, May 2024) | DeepSeek-V3.2-Speciale |
|---|---|---|
| Provider | DeepSeek | DeepSeek |
| Noometry Index | 39.7 | 39.1 |
| Rank | — | 159 |
| Context | — | 128K |
| Input $/M | — | $0.58 |
| Output $/M | — | $1.68 |
| Weights | Open | Open |

## Coding

- DeepSeek-V2 (MoE-236B, May 2024): 39.8 (#141)
- DeepSeek-V3.2-Speciale: 39.9 (#136)

| Benchmark | DeepSeek-V2 (MoE-236B, May 2024) | DeepSeek-V3.2-Speciale |
|---|---|---|
| WeirdML | — | 46.7% |
| BigCodeBench Instruct | 48.9% | — |
| BigCodeBench Complete | 59.4% | — |

## Reasoning

- DeepSeek-V2 (MoE-236B, May 2024): —
- DeepSeek-V3.2-Speciale: 32.1 (#65)

| Benchmark | DeepSeek-V2 (MoE-236B, May 2024) | DeepSeek-V3.2-Speciale |
|---|---|---|
| SimpleBench | — | 52.6% |
| BIG-Bench Hard | 78.8% | — |
| Epoch Capabilities Index | 124.78 | — |
| HellaSwag | 87.1% | — |
| PIQA | 83.9% | — |
| WinoGrande | 86.3% | — |

## Knowledge

- DeepSeek-V2 (MoE-236B, May 2024): —
- DeepSeek-V3.2-Speciale: —

| Benchmark | DeepSeek-V2 (MoE-236B, May 2024) | DeepSeek-V3.2-Speciale |
|---|---|---|
| ARC (AI2) Challenge | 92.2% | — |
| MMLU | 78.4% | — |
| TriviaQA | 80% | — |

## Writing & Preference

- DeepSeek-V2 (MoE-236B, May 2024): —
- DeepSeek-V3.2-Speciale: 46.1 (#217)

| Benchmark | DeepSeek-V2 (MoE-236B, May 2024) | DeepSeek-V3.2-Speciale |
|---|---|---|
| EQ-Bench Creative Writing | — | 1276 |

## FAQ

### Is DeepSeek-V2 (MoE-236B, May 2024) better than DeepSeek-V3.2-Speciale?

DeepSeek-V3.2-Speciale has enough public results to be ranked (#159); DeepSeek-V2 (MoE-236B, May 2024) does not yet, so treat this comparison as directional.

### Is DeepSeek-V2 (MoE-236B, May 2024) or DeepSeek-V3.2-Speciale better for coding?

They score almost the same on coding (39.8 vs 39.9); test both on your own repository before choosing.

### How many benchmarks do DeepSeek-V2 (MoE-236B, May 2024) and DeepSeek-V3.2-Speciale share?

0 benchmarks have published results for both models. DeepSeek-V2 (MoE-236B, May 2024) has 10 scored results on Noometry and DeepSeek-V3.2-Speciale has 3.