Question: Chose pipelines OR OpenAI, but never both?

[TomTom101]

When I want to use custom pipelines that run on my locally deployed "pipelines" instance (http://localhost:9099) I do set them up in "Manage OpenAI API Connections". This is somewhat counter intuitive, but my main pain with this is that I cannot just offer users arbitrary OpenAI models anymore.
How would I add, say, 5 different OpenAI models without having to create a own Pipeline for each model?

[drorm]

Create a new connection.
Enter https://api.openai.com/v1 and the key.
And you get all the openai models.

[TomTom101]

Thanks! On a wide screen, that small "+" icon is hardly visible and the screen in general is not overly intuitive imo.

I was thinking there is a single "OpenAI API" which you can (de)activate (for what ever reason) and set it to a single connection.

Aren't all these, including Ollama, just "OpenAI API compatible endpoints" of which you can have as many as you want?

Either way, now I got it, thanks again!

[drorm]

You're welcome.
Totally agreed about the '+' hardly being visible.