Use a Open-WebUI defined openai API URL in a pipeline?

[bartmch]

I have an existing OpenAI API Connection "http://model:8000/v1" which is the default chat model in my open-webui.
I would like to use this same "http://model:8000/v1" LLM inside the pipeline to e.g. do nl2sql. Maybe I missed it but I can only see ollama and other "externally defined" LLM's.
How can I use the open-webui defined OpenAI API URL and use it in my pipeline?