In this paper, we focus on two popular instances of greedy algorithms, namely orthogonal matching pursuit (OMP) and orthogonal least squares (OLS). The suboptimal nature of OMP and OLS has led many researchers to study conditions under which these procedures succeed in recovering the true sparse vector. This question has been widely addressed for OMP in the recent years, including worst-case uniform and probabilisticanalyses. Although OLS has been known in the literature for a few decades (under different names, exact recovery analysis for OLS have only appeared very recently. Most of the existing works deal with uniform guarantees: these conditions ensure the success of OMP/OLS for a given sparsity level (or a given support) irrespective of the magnitude of the non-zero coefficients. In contrast with these works, we derive new guarantees of success accounting for the decay of the non-zero elements of the sparse vector. Our conditions are expressed in terms of the mutual coherence of the dictionary $\mu$ and encompass, as particular cases, some well-known results of the literature.