We propose a novel step size strategy for the extragradient method to solve root-finding and min-max optimization problems under relaxed Lipschitz assumptions