In this article, a new maximum power point tracking algorithm based on a modified butterfly optimization algorithm has been proposed. The proposed method is capable of differentiating between different partial shading patterns, uniform shading, solar intensity, and load variation conditions with fast convergence speed (CS). Only one dynamic variable is used as a tuning parameter reducing the complexity of the algorithm. The search space skipping method has been proposed to improve the CS. The proposed method is hybridized with a constant impedance method to improve the response time of the system for fast varying load variations. The proposed method has been validated experimentally on the SEPIC converter topology with a sampling time of 0.05 s. The experimental validation proved the average tracking time for different shading patterns is less than 1 s with steady-state efficiency of 99.85% on average. The CS for uniform shading conditions is improved by 47.20%. The response to load variation is also improved by 86.15% and becomes eligible to be utilized for fast varying load variations. Finally, the comparison table based on the MPPT rating has been presented to determine the effectiveness of the proposed method among other popular metaheuristic approaches used for MPPT.