Grok 4.20 0309 v2 vs Llama Nemotron Super 49B v1.5
xAI vs NVIDIA — benchmarks, pricing, and capabilities side by side.
- •Grok 4.20 0309 v2 has the higher intelligence index (91.1 vs 79.4)
- •Llama Nemotron Super 49B v1.5 is cheaper ($0.10 vs $2.00 per 1M input)
- •Grok 4.20 0309 v2 is faster
| Grok 4.20 0309 v2 | Llama Nemotron Super 49B v1.5 | |
|---|---|---|
| Intelligence index | 91.1 | 79.4 |
| Developer | xAI | NVIDIA |
| Type | LLM | LLM |
| Access | — | — |
| Context window | — | — |
| Input price | $2.00 / 1M | $0.10 / 1M |
| Output price | $6.00 / 1M | $0.40 / 1M |
| Speed | 105 tok/s | 51 tok/s |
| Released | April 7, 2026 | July 25, 2025 |
| Parameters | — | — |
| Input modalities | — | — |
| Output modalities | — | — |
Shared benchmarks
Grok 4.20 0309 v2
Llama Nemotron Super 49B v1.5
GPQA Diamond
91.1
74.8
Humanity’s Last Exam
32.2
6.8