it depends on your cu oxidation number. how many electrons are in your cu atom?

Bascially, you should know the oxidation state of copper in your system. 

Then, check a number of electrons what you have interestted in. ( odd ? or even?)

Then you can calculate multiplicity with consdering a charge state.

you could try to apply finite temperature DFT, also called Fermi smearing, to determine the most stable occupation. But I am not sure, if this is available in your program. Or simply try out all the different multiplicities.

Mr.  Kai-yang,

Please pay attention to:

http://www.gaussian.com/g_tech/g_ur/m_input.htm

I would suggest performing the relaxation at a number of spin multiplicities and determining the ground state spin that way. Using Fermi smearing will NOT always allow the spin to change so as to find the ground state surface, though some people think that it will. You may want to try using the "stable=opt" command, which will in some cases automatically check to see if flipping a spin will lower the energy, though this is not fool proof either, in my experience. One thing to keep your eye on as well is the value of . Especially with the lower spin states, you may get spin contamination from higher spin states that are close in energy. Whether you consider this to be a big deal or not somewhat depends on your philosophy, but regardless you should always check it when different spin surfaces are close in energy. Good luck!