Head to head

Llama 3.1 8B Instant vs LFM2.5 8B-A1B

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.

Llama 3.1 8B InstantLFM2.5 8B-A1B
Provider Groq Together AI
Input / MTok $0.05 $0.03
Output / MTok $0.08 $0.12
Cache read / MTok
Context window 131,072 32,768
Blended 1M in + 3M out $0.29 $0.39
Status ga ga

Llama 3.1 8B Instant detail   LFM2.5 8B-A1B detail