It can, but it'll take more set up since there's a backend we have configured that does the actual LLM communication. You'd need to set that up internally for an on-prem deployment. Can you send us a message (https://www.lotas.ai/contact) if you want to talk more?