To have your own LLM provider, you have to run different software. Ollama is one option (though it also by default uses external provider) and llama.cpp is another. Jan and LMStudio are other options for actually running a model on your own hardware, to which you can direct OpenCode.
bigyabai•1h ago
ciwolex•1h ago