The Gustafson–Kessel (GK) algorithm, an extension of the fuzzy c-means (FCM) clustering method, effectively handles non-spherical clusters but struggles with uncertainty in membership assignments. To address this limitation, we propose the n-Pythagorean Fuzzy Gustafson–Kessel (n-PyGK) algorithm, which incorporates an inherent hesitation degree to enhance clustering performance. The proposed algorithm is evaluated on both synthetic and real-world datasets, including the Iris dataset, using nine clustering metrics. We analyze its behavior under varying parameter settings and compare its performance with traditional clustering algorithms. Experimental results demonstrate that n-PyGK offers improved clustering accuracy and greater flexibility in parameter selection, enabling optimal performance for specific clustering indices.