Looks like this also includes llama.cpp. You can either run everything local or connect it to a remote server as well.
