Head to head
LFM2.5 8B-A1B vs Llama 3.1 8B Instant
On a blended workload of one million input tokens and three million output tokens, Llama 3.1 8B Instant is 1.3× cheaper than LFM2.5 8B-A1B.
| LFM2.5 8B-A1B | Llama 3.1 8B Instant | |
|---|---|---|
| Provider | Together AI | Groq |
| Input / MTok | $0.03 | $0.05 |
| Output / MTok | $0.12 | $0.08 |
| Cache read / MTok | — | — |
| Context window | 32,768 | 131,072 |
| Blended 1M in + 3M out | $0.39 | $0.29 |
| Status | ga | ga |