I have no issues using it with llama.cpp.