This is for connecting local server models like llamacpp, ollama, etc... to LM studio. Any server with a base url. The original plugin does not allow the to add thier own model names (as far as I know) so I could not load/unload or query my models.

This plugin uses the user's /v1/models endpoint to list their available models, and adds them to the model dropdown in the UI, overwriting the OpenAI and Anthropic defaults.

To do this:

Install this plugin
Enter base URL
Type something in the chat box and send that prompt
- Your API endpoint will then populate the model dropdown with the models from your endpoint at /v1/models.
Uninstall this plugin
Reinstall this plugin Your dropdown will now be populated with your endpoint's model list

openai-compat-endpoint