Clustering is an essential task in data analysis with applications in various fields. Although the widely used K-means algorithm has limitations, several heuristic approaches have been proposed to improve its performance. One such algorithm is GSA-KM, which combines the gravitational search algorithm (GSA) with K-means (KM) clustering. However, GSA itself has certain drawbacks that can be addressed through parameter tuning and the integration of chaotic optimization techniques. The Chaotic k-best GSA (CKGSA) is a hybrid algorithm that utilizes chaos theory to enhance the precision, convergence rate, and global search capability of GSA. This paper proposes a new clustering approach that combines CKGSA and K-Means algorithms. Integrating CKGSA improves the search strategy, while K-Means contributes to effective clustering results. Our proposed algorithm shows promise for enhancing clustering performance in terms of both efficiency and quality. This paper provides a comprehensive overview of the methodology, detailed experimental results, and a comprehensive discussion of the findings. We have used three benchmark datasets (Iris, Wine, and Glass) for the comparative analysis purpose, where the proposed approach shows better performance in terms of intra-cluster distance, standard deviation and average execution times.