Head to head

GPT-OSS 20B vs LFM2.5 8B-A1B

On a blended workload of one million input tokens and three million output tokens, LFM2.5 8B-A1B is 1.7× cheaper than GPT-OSS 20B.

GPT-OSS 20BLFM2.5 8B-A1B
Provider Together AI Together AI
Input / MTok $0.05 $0.03
Output / MTok $0.2 $0.12
Cache read / MTok
Context window 128,000 32,768
Blended 1M in + 3M out $0.65 $0.39
Status ga ga

GPT-OSS 20B detail   LFM2.5 8B-A1B detail