Comment by taytus

At current prices, and considering these OS Models' performance, investing in local inference sounds like a bad idea.