Nowadays, the Permanent Magnet Synchronous Motors (PMSM) are gaining popularity among electric motors due to their high efficiency, high-speed operation, ruggedness, and small size. PMSM motors comprise a trapezoidal electromotive force which is also called synchronous motors. Direct Torque Control (DTC) has been extensively applied in speed regulation systems due to its better dynamic behaviour. The controller manages the amplitude of torque and stator flux directly using the direct axis current. To manage the motor speed, the torque error, flux error, and projected location of flux linkage are employed to adjust the switching sequence of inverter. One of the most common problems encountered in a PMSM motor is Torque ripple, which is recreated by power electronic commutation and a better controller reduces the ripples to increase the drive's performance. Here, DTC control of PMSM is controlled by Golden Eagle Optimization (GEO) optimized Adaptive Neuro-Fuzzy Inference System (ANFIS). Learning parameters of the ANFIS are optimized for various operating functions by GEO in terms of speed and torque. In this work, simulation results are carried out in MATLAB, it shows that GEO-ANFIS controller is used in conjunction with a PMSM motor to attain less torque ripple up to 0.44 Nm and maintain the speed with a distortion error of 2.12 % when compared with Space Vector Pulse Width Modulation based DTC (SVPWM-DTC) and Dual Cost Function Model Predictive Direct Speed Control (DCF-MPDSC).