Simulation of grinding is a topic of great interest due to the wide application of the process in modern industry. Several modeling methods have been utilized in order to accurately describe the complex phenomena taking place during the process, the most common being the Finite Element Method (FEM) and the Artificial Neural Networks (ANN). In the present work, a FEM model and an ANN model for precision surface grinding, are presented. Furthermore, a new approach, a combination of the aforementioned methods, is proposed, and a hybrid model is presented. This model comprises the advantages of both FEM and ANN models. The three kinds of models described in this work are able to accurately predict several grinding features that define the outcome of the process and the quality of the final product.