DeepSeek V3.2 Speciale vs Devstral 2
DeepSeek vs Mistral AI — benchmarks, pricing, and capabilities side by side.
- •DeepSeek V3.2 Speciale has the higher intelligence index (89.9 vs 54.3)
- •DeepSeek V3.2 Speciale is cheaper ($0.29 vs $0.40 per 1M input)
- •Devstral 2 has a larger context window (262K)
| DeepSeek V3.2 Speciale | Devstral 2 | |
|---|---|---|
| Intelligence index | 89.9 | 54.3 |
| Developer | DeepSeek | Mistral AI |
| Type | LLM | LLM |
| Access | Open weights | Open weights |
| Context window | 163,840 tokens | 262,144 tokens |
| Input price | $0.29 / 1M | $0.40 / 1M |
| Output price | $0.43 / 1M | $2.00 / 1M |
| Speed | — | 51 tok/s |
| Released | December 1, 2025 | December 9, 2025 |
| Parameters | — | — |
| Input modalities | Text | Text |
| Output modalities | Text | Text |
Shared benchmarks
DeepSeek V3.2 Speciale
Devstral 2
AIME 2025
96.7
36.7
GPQA Diamond
87.1
59.4
Humanity’s Last Exam
26.1
3.6
LiveCodeBench
89.6
44.8
MMLU-Pro
86.3
76.2