here's an website for a community-ran db on LLM models with details on configs for their token/s: https://inferbench.com/