Llama Guard 4 12 B vs Us Gov Nvidia Nemotron Super 3 120 B: cost comparison
Prices updated 2026-10-28 · computed from live provider data
Short answer: Llama Guard 4 12 B is cheaper on a
3:1 input/output blend — $0.180 per million tokens versus
$0.330, a 45% difference.
Us Gov Nvidia Nemotron Super 3 120 B offers the larger context window at 256,000 tokens, against 163,840 for Llama Guard 4 12 B.
Head to head
| Llama Guard 4 12 B | Us Gov Nvidia Nemotron Super 3 120 B | |
|---|---|---|
| Input, per 1M tokens | $0.180 | $0.180 |
| Output, per 1M tokens | $0.180 | $0.780 |
| Blended 3:1, per 1M | $0.180 | $0.330 |
| Context window | 163,840 | 256,000 |
| Vendor | meta-llama | us |
Monthly cost at real volumes
| Workload (input / output) | Llama Guard 4 12 B | Us Gov Nvidia Nemotron Super 3 120 B |
|---|---|---|
| 1M in / 250K out | $0.225 | $0.375 |
| 10M in / 2.5M out | $2.25 | $3.75 |
| 100M in / 25M out | $22.50 | $37.50 |
Which should you pick?
Pick Us Gov Nvidia Nemotron Super 3 120 B when you need the longer context window (256,000 vs 163,840 tokens). Pick Llama Guard 4 12 B when price per token dominates your workload — high-volume classification, extraction, or batch summarisation.