C_R–M–z relation of the eRASS1 clusters and the best-fit model in Eq. (22). Its fully marginalized uncertainties are shown as grey shaded regions. The left panel shows the mass trend normalized at the pivotal redshift z_piv = 0.35 after accounting for the redshift-dependent factors, while the right panel similarly presents the redshift scaling normalized at the pivotal mass $M_{\mathrm{piv}}=1.4\times {10}^{14}{h}^{-1}{\mathrm{M}}_{\odot }$. In both panels, the eRASS1 clusters are colour-coded based on the redshift (left panel) and posterior-sampled halo mass (right panel). For comparison, the results from Grandis et al. (2024, green dash-dotted lines) and Kleinebreil et al. (2025, purple lines), which are based on the weak-lensing mass calibration from the DES and KiDS surveys, respectively, are also plotted. The observed C_R–M–z relation reveals a mass trend that is steeper than the self-similar prediction (red dashed line), while its redshift trend remains statistically consistent with the self-similar scaling. The normalization of the self-similar predictions (red dashed lines) is fixed to the best-fit A_X.