The mixed layer plays a crucial role as an entry or exit point for heat, salt, momentum and nutrients from the surface to the deep ocean. We propose here a generic framework to evaluate the mixed layer depth (MLD) dynamics across a wide range of forcings and preconditioning conditions. To do so, we propose to use a physically-based parameter space formed by two dimensionless numbers: $\lambda_s$ the relative contribution of the buoyancy flux and the wind in the surface layer, and the Richardson number $R_h$ which characterizes the stability of the water column at the mixed layer base. Four MLD dynamics regimes (“restratification”, “stable”, “deepening” and “strong deepening”) are defined based on the values of the normalized temporal evolution of the MLD. We show the robustness and the predictive skill of the parameter space in a context of 1D simulations by showing that to a given ($\lambda_s$, $R_h$) corresponds a predictable MLD dynamics regime. Finally, we present two applications showing how the parameter space can be used with 3D ocean realistic models. We discuss the impact of the horizontal resolution (1°, 1/12° or 1/60°) and the Gent McWilliams parameterization on the MLD dynamics regimes.