So, you go to: kobold github
and download any model from: collab (scroll down to the links.

Make sure you go to the files and versions page and right click and copy file address of the download button or press the download button

Open koboldcpp.exe normally or in admin
go to GGUF text model and click browse or paste in the link adress
For browse, click on the downloaded model

make sure to set the context size to 16000-36000
make sure you have atleast 8k-32k memory depending on what settings you're running on.
make sure you have a good gpu

then click launch
Once it says: Note: For third party Ollama API Emulation, you should set the port to 11434.

Starting Kobold API on port 5001 at http://localhost:5001/api/

Starting OpenAI Compatible API on port 5001 at http://localhost:5001/v1/

Starting llama.cpp secondary WebUI at http://localhost:5001/lcpp/
copy http://localhost:5001/lcpp/v1/

put it into proxy
edit a configuration, then paste it into the proxy url, click on /chat/completions, you can add anything to the other config settings of it, make sure you dont change the proxy url, api key is not needed.
press save changes, then save settings, then refresh

Open the api configs, then click test on the configuration.
then click save settings.