OpenAI API LiteLLM proxy server

[mkagit]

Hi!
Since I'm already running a LiteLLM proxy server for another project I'm wondering if I could use it also for the llmcord. I use it with virtual keys for tracking usage, set budget, guardrails and to attach a group of models to a team virtual key. I manage all from the Admin UI.
Is it possible to set the baseUrl as the LiteLLM OpenAI API compatible endpoint? Would be like this: http://myserver.com:4000/v1/chat/completions

[jakobdylanc]

Yes this should be possible. Don't include the /chat/completions part. Just add this new entry to providers in your config.yaml:

providers:
  litellm:
    base_url: http://myserver.com:4000/v1

And then set model to litellm/<model name>.