DeepSeek V3.2 Speciale vs Llama Nemotron Super 49B v1.5
DeepSeek vs NVIDIA — benchmarks, pricing, and capabilities side by side.
- •DeepSeek V3.2 Speciale has the higher intelligence index (89.9 vs 79.4)
- •Llama Nemotron Super 49B v1.5 is cheaper ($0.10 vs $0.29 per 1M input)
| DeepSeek V3.2 Speciale | Llama Nemotron Super 49B v1.5 | |
|---|---|---|
| Intelligence index | 89.9 | 79.4 |
| Developer | DeepSeek | NVIDIA |
| 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 | — | 51 tok/s |
| Released | December 1, 2025 | July 25, 2025 |
| Parameters | — | — |
| Input modalities | Text | — |
| Output modalities | Text | — |
Shared benchmarks
DeepSeek V3.2 Speciale
Llama Nemotron Super 49B v1.5
AIME 2025
96.7
76.7
GPQA Diamond
87.1
74.8
Humanity’s Last Exam
26.1
6.8
LiveCodeBench
89.6
73.7
MMLU-Pro
86.3
81.4