Request to add support for locally hosted ollama

[drmetro09]

It would be very helpful if we could use local hosted models with ollama

[KalimaSan]

Same here, configuring local LLMs will unlock a lot of more users I think. It would be awesome if an env variable config could be added.