Summary Model predictive control has been introduced as a robust and high‐performance control method for the permanent‐magnet synchronous motor (PMSM) due to its accuracy, simplicity, and flexibility in the controlled variables. In this article, an optimized finite control set‐model predictive control method (FCS‐MPC) for the surface mounted and interior PMSM supplied by a 2‐level voltage source inverter (2L‐VSI) is proposed. In each sampling time of conventional FCS‐MPC, voltage is implemented only one time, which increases the motor's steady‐state errors. In order to decrease such errors, recently published papers have proposed combining two or three voltage vectors in a sampling time. Although using several combinations of voltage vectors increases the complexity, it leads to superior performance. In this article, a more desirable result is obtained by using combinations of five voltage vectors. In the proposed method, two extra candidates, in addition to the common choices, are added to each period, which leads to generating a voltage vector closer to the reference one and thereby, resulting in better performance. Moreover, the principle of dead‐beat control is used to accelerate the calculation; also proposed method calculates the duty cycle just for the selected vector instead of all vectors, so execution time decreases. The simulation results obtained by MATLAB/Simulink confirm the efficacy of the proposed MPC method.