To the best of my knowledge, one of the widely-used software based on effective Hamiltonian was developed by M. F. Reid. The other one was written by Y. Y. Yeung.

At the level of first principle, MolCas/MolPro is suitable.

Dear Zeltzin Reyes,

everything depends on your ultimate goal. In other words, what do you need the erbium energy levels for? In some cases, you can be limited to semiempirical methods, and, in other cases, you need to apply only fist-principal calculations.