For the chat completions endpoint, Edgen supports any model on Hugging Face that is tagged with the library GGUF. You can explore the available models in the HuggingFace repository.
For the audio transcriptions endpoint, Edgen supports all whisper.cpp models:
|Whisper (all variants)
Just change the configuration file! Check Documentation » Configuration and if Edgen cannot find the model you specified locally, it'll download it automatically from HuggingFace. See also API Reference » Models.
You can also download your model manually and copy it to the model directory. In this case, Edgen will not manage this model.