Enabling multiple OpenAI API-compatible models

[gwtaylor]

Feature request

It would be nice to be able to include multiple OpenAI API-compatible models, for example, one hosted by OpenAI and another hosted by Azure (GitHub Models). To my knowledge, it is only possible to consider a single OpenAI API-compatible model, though the OPENAI.url can be set to non-OpenAI hosts (like Azure).

Why?

The Ollama (local) configuration permits multiple models. There are many diverse OpenAI API-compatible models that may provide different commit messages and users may want to consider those outputs.

Alternatives

No response

Additional context

No response

[tak-bro]

I understand your request. It would be good to support multiple models simultaneously by extending the OpenAI based code..
I'll consider implementing support for multiple models using custom URL configurations (CUSTOM.url, CUSTOM2.url, etc.). Thank you for the feature request.