DeepSeek V3.2 Speciale vs Llama 3.1 8B Instruct
DeepSeek vs Meta — benchmarks, pricing, and capabilities side by side.
- •DeepSeek V3.2 Speciale has the higher intelligence index (89.9 vs 45)
- •Llama 3.1 8B Instruct is cheaper ($0.02 vs $0.29 per 1M input)
- •DeepSeek V3.2 Speciale has a larger context window (164K)
| DeepSeek V3.2 Speciale | Llama 3.1 8B Instruct | |
|---|---|---|
| Intelligence index | 89.9 | 45 |
| Developer | DeepSeek | Meta |
| Type | LLM | LLM |
| Access | Open weights | Open weights |
| Context window | 163,840 tokens | 131,072 tokens |
| Input price | $0.29 / 1M | $0.02 / 1M |
| Output price | $0.43 / 1M | $0.05 / 1M |
| Speed | — | 2047 tok/s |
| Released | December 1, 2025 | July 23, 2024 |
| Parameters | — | — |
| Input modalities | Text | Text |
| Output modalities | Text | Text |
Shared benchmarks
DeepSeek V3.2 Speciale
Llama 3.1 8B Instruct
AIME 2025
96.7
4.3
GPQA Diamond
87.1
30.4
Humanity’s Last Exam
26.1
5.1
LiveCodeBench
89.6
11.6
MMLU-Pro
86.3
48.3