The standard library does have some specialisation internally for certain iterators and collection combinations. Not sure if it will optimise that one specifically, but Vec::into_iter().collect::<Vec>() is optimised (it may look silly, but it comes up with functions returning impl Iterator