Head to head
Llama Prompt Guard 2 22M vs Ministral 3 8B
On a blended workload of one million input tokens and three million output tokens, Llama Prompt Guard 2 22M is 5.0× cheaper than Ministral 3 8B.
| Llama Prompt Guard 2 22M | Ministral 3 8B | |
|---|---|---|
| Provider | Groq | Mistral |
| Input / MTok | $0.03 | $0.15 |
| Output / MTok | $0.03 | $0.15 |
| Cache read / MTok | — | — |
| Context window | — | — |
| Blended 1M in + 3M out | $0.12 | $0.60 |
| Status | preview | ga |