DeepSeek V3.2 Speciale vs Hermes 4 - Llama-3.1 70B
DeepSeek vs Nous Research — benchmarks, pricing, and capabilities side by side.
- •DeepSeek V3.2 Speciale has the higher intelligence index (89.9 vs 71.3)
- •Hermes 4 - Llama-3.1 70B is cheaper ($0.10 vs $0.29 per 1M input)
| DeepSeek V3.2 Speciale | Hermes 4 - Llama-3.1 70B | |
|---|---|---|
| Intelligence index | 89.9 | 71.3 |
| Developer | DeepSeek | Nous Research |
| Type | LLM | LLM |
| Access | Open weights | — |
| Context window | 163,840 tokens | — |
| Input price | $0.29 / 1M | $0.10 / 1M |
| Output price | $0.43 / 1M | $0.40 / 1M |
| Speed | — | 60 tok/s |
| Released | December 1, 2025 | August 27, 2025 |
| Parameters | — | — |
| Input modalities | Text | — |
| Output modalities | Text | — |
Shared benchmarks
DeepSeek V3.2 Speciale
Hermes 4 - Llama-3.1 70B
AIME 2025
96.7
68.7
GPQA Diamond
87.1
69.9
Humanity’s Last Exam
26.1
7.9
LiveCodeBench
89.6
65.3
MMLU-Pro
86.3
81.1