Average BLS runtime as a function of the maximum searched period, for a constant number of evenly spaced data points (10⁵). Three algorithms are compared: brute-force AstroPy BLS (red), FFA-based GERBLS (green), and GERBLS with automatic data downsampling (blue). The relative speed gain from using an FFA-based algorithm diminishes for longer orbital periods, unless the data are progressively downsampled at these periods. In all cases, GERBLS still substantially outperforms AstroPy in average runtime, even when the maximum search period is half as long as the total time baseline of the data (here, at P_max = 300 days).