Statistical criteria have long been the standard for selecting the best model for phylogenetic reconstruction and downstream statistical inference. While model selection is regarded as a fundamental step in phylogenetics, existing methods for this task consume computational resources for long processing time, they are not always feasible, and sometimes depend on preliminary assumptions which do not hold for sequence data. Moreover, while these methods are dedicated to revealing the processes that underlie the sequence data, in most cases they do not produce the most accurate trees. Notably, phylogeny reconstruction consists of two related tasks, topology reconstruction and branch-length estimation. It was previously shown that in many cases the most complex model, GTR+I+G, leads to topologies that are as accurate as using existing model selection criteria, but overestimates branch lengths. Here, we present ModelTeller, a computational methodology for phylogenetic model selection, devised within the machine-learning framework, optimized to predict the most accurate model for branch-length estimation accuracy. ModelTeller relies on a readily implemented machine-learning model and thus the prediction according to features extracted from the sequence data results in a substantial decrease in running time compared to existing strategies. We show that on datasets simulated under simple homogenous substitution models ModelTeller leads to branch-length estimation that is as accurate as the statistical model selection criteria. We then demonstrate that ModelTeller outperforms these criteria when more intricate patterns -that aim at mimicking realistic processes -are considered.