lib_mlx_load_model method
Loads a local model path and returns a JSON result: {"ok":true,"handle":1,"status":"ready","model_id":"..."}.
The v1 inference path is HTTP-only. FFI is reserved for model/server lifecycle management and returns retained JSON strings that callers must free with lib_mlx_free.
Implementation
ffi.Pointer<ffi.Char> lib_mlx_load_model(ffi.Pointer<ffi.Char> config_json) {
return _lib_mlx_load_model(config_json);
}