cortex.llamacpp is a high-efficiency C++ inference engine for edge computing. It is a dynamic library that can be loaded by any server at runtime.