This is one of the things Ada does really well. The standard says that the range of Float is implementation defined, but a program can reference the 'First and 'Last attributes of the Float type to determine the actual bounds for the compiler and runtime.
Better title: “What is the range of a number type for types recognized by C++”?
The range of numbers represented by an IEEE-754 floating-point is discontinous on the number line at any given precision. So there are a variety of possible answers with varying usefulness.