Despite a large amount of work being devoted to study the phase separation in solids, the underlying physical mechanism responsible for such diffusive first-order phase transitions remains difficult to model outside the spinodal regime, i.e., in the nucleation and growth regime. This work presents an alternative of the classical nucleation theory for modeling phase separation in this regime, even for systems far from the solubility limit, i.e., for high degree of meta-stability where the classical nucleation theory does not hold. This method then allows a direct comparison between simulations and experiments always performed in solids with a high degree of meta-stability.