The objective of this research study was to propose a modified Jelly Fish Optimizer (JFO) for permanent magnet synchronous motors (PMSMs) to reduce torque ripple. Torque ripple is a common problem in PMSMs that can cause mechanical vibrations and noise, reduce the system's efficiency, and shorten the motor's lifespan. Therefore, reducing torque ripple is critical to improving the performance and reliability of PMSMs. The jellyfish algorithm is a metaheuristic optimization method inspired by the behavior of jellyfish in the ocean. It is a population-based algorithm that employs a search strategy based on the collective motion of the jellyfish swarm. In this study, the jellyfish algorithm was modified to optimize a torque ripple minimization objective function for the PMSM system. The objective function was designed based on the Fourier analysis of the torque waveform and aimed to reduce the harmonics in the torque ripple spectrum. Here this proposes Jelly fish algorithm based artificial neuro fuzzy interference system (ANFIS) Direct torque control (DTC). ANFIS parameters are optimized using Jelly fish functions. It gives less torque ripple and improved performance compared to other conventional controller. Simulation is done on PMSM model using MATLAB/Simulink. This algorithm gives a less torque of 0.3 Nm compared to Space vector pulse width modulation based Direct Speed Control (SVPWM-DTC).