Is your feature request related to a problem? Please describe.
When adding a custom OpenAI compatible provider, the available models have to be added manually. The list can be retrieved using the /v1/models endpoint.
Describe the solution you'd like
For a custom OpenAI compatible provider, query the models and pre-show them in the list of available models in the UI, highlighted in a way to communicate that they are available but not configured locally yet. A user can save one and fill in the details like context size and capabilities, but you don't have to look up the model id manually. This would make quick integration of custom providers a lot more convenient.
Describe alternatives you've considered
Additional context