# DeepSeek-V3.2-Speciale vs Magistral Medium

> DeepSeek-V3.2-Speciale is the stronger model overall, scoring 39.1 to 34.8 on the Noometry Index.

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

## Summary

- The widest gap is in reasoning, where DeepSeek-V3.2-Speciale leads 32.1 to 8.7.
- DeepSeek-V3.2-Speciale is cheaper at $0.58 / $1.68 per million input/output tokens, against $2 / $5 for Magistral Medium.

## Snapshot

| | DeepSeek-V3.2-Speciale | Magistral Medium |
|---|---|---|
| Provider | DeepSeek | Mistral AI |
| Noometry Index | 39.1 | 34.8 |
| Rank | 159 | 224 |
| Context | 128K | 128K |
| Input $/M | $0.58 | $2 |
| Output $/M | $1.68 | $5 |
| Weights | Open | Open |

## Coding

- DeepSeek-V3.2-Speciale: 39.9 (#136)
- Magistral Medium: 38.6 (#161)

| Benchmark | DeepSeek-V3.2-Speciale | Magistral Medium |
|---|---|---|
| SciCode | — | 39.2% |
| WeirdML | 46.7% | — |
| LMArena Coding | — | 1319 |

## Reasoning

- DeepSeek-V3.2-Speciale: 32.1 (#65)
- Magistral Medium: 8.7 (#342)

| Benchmark | DeepSeek-V3.2-Speciale | Magistral Medium |
|---|---|---|
| ARC-AGI-2 | — | 0% |
| SimpleBench | 52.6% | — |
| Kagi LLM Benchmark | — | 16.2% |
| ARC-AGI-1 | — | 6.1% |
| CritPt | — | 0.3% |
| LMArena Hard Prompts | — | 1267 |

## Math

- DeepSeek-V3.2-Speciale: —
- Magistral Medium: 32.4 (#187)

| Benchmark | DeepSeek-V3.2-Speciale | Magistral Medium |
|---|---|---|
| LMArena Math | — | 1249 |

## Knowledge

- DeepSeek-V3.2-Speciale: —
- Magistral Medium: 32.9 (#199)

| Benchmark | DeepSeek-V3.2-Speciale | Magistral Medium |
|---|---|---|
| LMArena Expert | — | 1222 |

## Multilingual

- DeepSeek-V3.2-Speciale: —
- Magistral Medium: 39.8 (#219)

| Benchmark | DeepSeek-V3.2-Speciale | Magistral Medium |
|---|---|---|
| LMArena Non-English | — | 1232 |
| LMArena Chinese | — | 1225 |
| LMArena French | — | 1268 |
| LMArena German | — | 1248 |
| LMArena Japanese | — | 1176 |
| LMArena Korean | — | 1125 |
| LMArena Russian | — | 1224 |
| LMArena Spanish | — | 1269 |

## Instruction Following

- DeepSeek-V3.2-Speciale: —
- Magistral Medium: 66.2 (#207)

| Benchmark | DeepSeek-V3.2-Speciale | Magistral Medium |
|---|---|---|
| LMArena Instruction Following | — | 1254 |

## Long Context

- DeepSeek-V3.2-Speciale: —
- Magistral Medium: 39.3 (#178)

| Benchmark | DeepSeek-V3.2-Speciale | Magistral Medium |
|---|---|---|
| LMArena Longer Query | — | 1295 |

## Writing & Preference

- DeepSeek-V3.2-Speciale: 46.1 (#217)
- Magistral Medium: 46.5 (#214)

| Benchmark | DeepSeek-V3.2-Speciale | Magistral Medium |
|---|---|---|
| LMArena Text | — | 1255 |
| LMArena Creative Writing | — | 1245 |
| EQ-Bench Creative Writing | 1276 | — |
| LMArena Multi-Turn | — | 1275 |

## FAQ

### Is DeepSeek-V3.2-Speciale better than Magistral Medium?

DeepSeek-V3.2-Speciale is the stronger model overall, scoring 39.1 to 34.8 on the Noometry Index.

### Which is cheaper, DeepSeek-V3.2-Speciale or Magistral Medium?

DeepSeek-V3.2-Speciale is cheaper. It lists at $0.58 per million input tokens and $1.68 per million output tokens; Magistral Medium lists at $2 and $5.

### Is DeepSeek-V3.2-Speciale or Magistral Medium better for coding?

DeepSeek-V3.2-Speciale scores higher on coding benchmarks: 39.9 versus 38.6 in the Noometry coding category.

### Which has the bigger context window?

Both accept 128K tokens.

### How many benchmarks do DeepSeek-V3.2-Speciale and Magistral Medium share?

0 benchmarks have published results for both models. DeepSeek-V3.2-Speciale has 3 scored results on Noometry and Magistral Medium has 22.