Go to http://www.scattport.org/ and download the Mie code that is more suitable for you. Then go to http://refractiveindex.info/ and download the needed dielectric constants (or complex refractive index if you will)
As I understand it from reading Kriebeg, Mie scattering fails at 10nm and smaller. I've been using the Drude model, and incorporating the intraband transitions. But I haven only been able to find the Drude parameters for the dielectric and the dielectric values for bulk, for gold.
It is hard to establish at which size Mie theory fails. if i were you I would employ the following procedure:
1) Employ tabulated dielectric functions from the literature. You can find them at the web site I gave you in the first answer.
2) Correct the dielectric functions for the particle size: if I am not mistaken, the procedure is described in the Kreibig's book. The parameters to correct the dielectric functions are found in the Ashcroft and Mermin book.
3) Do your calculations: I would say that down to 10nm in diameter Mie theory is still ok. For lower diameters is tricky, I'd say that you could try down to 5nm, at least to get a qualitative idea, while for lower diameters is probably not going to work at all
I used Mie theory for gold nanoparticles down to 9.7 nm. I used measured values for the refractive index from Palik, Handbook of Optical Constants of Solids.