 The study proposes a generalized maximum likelihood approach to compare weighted financial network reconstruction methods using conditional entropy maximization. The method can be applied to any binary topology reconstruction method and uses available partial information to infer link weights. The results show that the most reliable method is obtained by addressing the best performing binary method with an exponential distribution of link weights and two safe variants are proposed. CREM underscore B is recommended for full uncertainty about the network topology or when the existence of some links is certain, as it is faster and reproduces empirical networks with highest generalized likelihood among the considered competing models. This article was authored by Federica Parisi, Tiziano Scottini, and Diego Goliskeli.