It depends on your transfection plan. If you're putting the Cas9 gene into the nucleus, then you'll need an NLS so that the protein can go into the nucleus after it is translated.
If you're putting the Cas9 protein into the cell, then you could use electroporation to open holes in both the plasma membrane and the nuclear envelope that the protein can go through.
I'd probably recommend adding an NLS to Cas9 just to make nuclear import more foolproof, unless you're concerned about the NLS peptide affecting the function of the Cas protein.