DeepHermes 3 - Llama-3.1 8B vs DeepSeek V3.2 Speciale
Nous Research vs DeepSeek — benchmarks, pricing, and capabilities side by side.
- •DeepSeek V3.2 Speciale has the higher intelligence index (89.9 vs 23.5)
- •DeepHermes 3 - Llama-3.1 8B is cheaper ($0.00 vs $0.29 per 1M input)
| DeepHermes 3 - Llama-3.1 8B | DeepSeek V3.2 Speciale | |
|---|---|---|
| Intelligence index | 23.5 | 89.9 |
| Developer | Nous Research | DeepSeek |
| Type | LLM | LLM |
| Access | — | Open weights |
| Context window | — | 163,840 tokens |
| Input price | $0.00 / 1M | $0.29 / 1M |
| Output price | $0.00 / 1M | $0.43 / 1M |
| Speed | — | — |
| Released | February 13, 2025 | December 1, 2025 |
| Parameters | — | — |
| Input modalities | — | Text |
| Output modalities | — | Text |
Shared benchmarks
DeepHermes 3 - Llama-3.1 8B
DeepSeek V3.2 Speciale
GPQA Diamond
27
87.1
Humanity’s Last Exam
4.3
26.1
LiveCodeBench
8.5
89.6
MMLU-Pro
36.5
86.3