The distinction between "compiled" and "interpreted" languages is mostly artificial these days.

People tend to call them "compiler" if they store the byte code somewhere on disk, and you can easily (for whatever values of "easily") execute this byte code. Otherwise they're called "interpreter"

But usually the process from source file to machine code involves both compilation (from abstract syntax trees to some kind of opcode tree) and interpretation (for example for optimization).

For example most compilers would optimize a= 2 * 5 to a = 10, which is an interpretation at compile time.