The paper introduces a new software tool called "Smart Model Selection" (SMS) for phylogenetic analysis, which is designed to efficiently select the best substitution model and rate-heterogeneity model (RAS) using likelihood-based criteria such as AIC and BIC. SMS is integrated into the PhyML web server and can also be used as a standalone application. The tool employs heuristic strategies to avoid extensive model testing, significantly reducing computational time while maintaining or improving model selection accuracy compared to exhaustive methods. For proteins, SMS considers 17 substitution matrices and two RAS options, computing likelihood values for approximately 30 models on average. For DNA, it uses four substitution matrices and five RAS options, computing likelihood values for about 6-7.5 models on average. SMS outperforms jModelTest2 and ProtTest in terms of speed and model selection, making it a valuable tool for large-scale phylogenetic analyses.The paper introduces a new software tool called "Smart Model Selection" (SMS) for phylogenetic analysis, which is designed to efficiently select the best substitution model and rate-heterogeneity model (RAS) using likelihood-based criteria such as AIC and BIC. SMS is integrated into the PhyML web server and can also be used as a standalone application. The tool employs heuristic strategies to avoid extensive model testing, significantly reducing computational time while maintaining or improving model selection accuracy compared to exhaustive methods. For proteins, SMS considers 17 substitution matrices and two RAS options, computing likelihood values for approximately 30 models on average. For DNA, it uses four substitution matrices and five RAS options, computing likelihood values for about 6-7.5 models on average. SMS outperforms jModelTest2 and ProtTest in terms of speed and model selection, making it a valuable tool for large-scale phylogenetic analyses.