Accept Lm Studio api

Instead of using its own Olama, I would love to be able to use my Own LLM in an OpenAI-compatible API (in my case, LM Studio). This allows me to reduce space consumption on my machine, avoiding duplicate LLM files and unnecessary processes.