DeepSeek V3.2 Speciale vs Phi 4
DeepSeek vs Microsoft — benchmarks, pricing, and capabilities side by side.
- •DeepSeek V3.2 Speciale has the higher intelligence index (89.9 vs 47.6)
- •Phi 4 is cheaper ($0.07 vs $0.29 per 1M input)
- •DeepSeek V3.2 Speciale has a larger context window (164K)
| DeepSeek V3.2 Speciale | Phi 4 | |
|---|---|---|
| Intelligence index | 89.9 | 47.6 |
| Developer | DeepSeek | Microsoft |
| Type | LLM | LLM |
| Access | Open weights | Open weights |
| Context window | 163,840 tokens | 16,384 tokens |
| Input price | $0.29 / 1M | $0.07 / 1M |
| Output price | $0.43 / 1M | $0.14 / 1M |
| Speed | — | 33 tok/s |
| Released | December 1, 2025 | January 10, 2025 |
| Parameters | — | — |
| Input modalities | Text | Text |
| Output modalities | Text | Text |
Shared benchmarks
DeepSeek V3.2 Speciale
Phi 4
AIME 2025
96.7
18
GPQA Diamond
87.1
56.1
Humanity’s Last Exam
26.1
4.1
LiveCodeBench
89.6
23.1
MMLU-Pro
86.3
70.4