C++23 added several fixed width floating-point type aliases defined in <stdfloat> including float32_t. Note that float32_t is only guaranteed to exist in an implementation if the pre-preprocessor macro __STDCPP_FLOAT32_T__ is defined as 1.