How do i add a self hosted Ollama Gemma4 model running on another machine?

[lpepen-re]

hello, i could use some help setting up claurst on my test laptop. i have a mac mini running ollama and serving gemma 4 e4b. in the openclaude.json i added the local ip of my computer in the providers section but claurst is unable to connect to it. I was successfully able to connect OpenCode to it, but claurst keeps thinking that ollama is running locally.

i saw #86 issue was opened, which is very relevant. i dont want to flood that issues thread so im opening this discussion thread

[AdamBajger]

if you are experiencing issues, or feel lost and would appreciate some docs to help you, definitely open an issue. The discussion here should be for the more open-ended questions, such as where to take this project further. I cannot help you at this moment to setup ollama, as I am not using it myself, but there should be a a settings.json file in the ~/.claurst directory, where you can configure providers. Try to fill in your ollama provider with custom "api_base", if that works. Otherwise, I am afraid I will reccomend you to upvode that issue, watch it closely and maybe provide some end-user feedback for us to resolve the issue more thoroughly. :)