This paper presents a bat algorithm variant, the Hybrid Bat Algorithm (HBA) for data clustering. HBA is a hybrid between a multi-modal variant of the bat algorithm (BA) and the K-Means algorithm. The multi-modal bat variant addresses two concerns regarding BA. It elevates the local convergence issue of BA by moving the bats toward their best neighbor instead of the global best solution. It further improves the lost convergence speed issue by incorporating a parameter that refines the random movement of a bat around itself. HBA employs K-Means with the multi-modal variant to further enhance its exploration and exploitation capabilities. Additionally, once every generation, HBA employs K-Means to improve the global best solution. Ten real-life datasets have been employed to compare HBA to MMBAIS, K-Means, and ten other contemporary nature-inspired clustering algorithms. The encouraging outcomes highlight the effectiveness of HBA.