The article introduces CAFE (Computational Analysis of gene Family Evolution), a computational tool designed to statistically analyze the evolution of gene family sizes over a phylogenetic tree. CAFE models gene family size changes using a stochastic birth and death process, allowing for the estimation of global birth and death rates, inference of ancestral gene family sizes, and identification of gene families with accelerated gain and loss rates. The tool can be run through a graphical interface or command-line version, and it provides methods to identify significant branches where evolutionary changes have occurred. CAFE is implemented in Java and is available for Mac OSX, Windows, and Linux systems. The authors discuss the computational aspects, including the use of Monte Carlo sampling and caching of probabilities to handle large gene families efficiently.The article introduces CAFE (Computational Analysis of gene Family Evolution), a computational tool designed to statistically analyze the evolution of gene family sizes over a phylogenetic tree. CAFE models gene family size changes using a stochastic birth and death process, allowing for the estimation of global birth and death rates, inference of ancestral gene family sizes, and identification of gene families with accelerated gain and loss rates. The tool can be run through a graphical interface or command-line version, and it provides methods to identify significant branches where evolutionary changes have occurred. CAFE is implemented in Java and is available for Mac OSX, Windows, and Linux systems. The authors discuss the computational aspects, including the use of Monte Carlo sampling and caching of probabilities to handle large gene families efficiently.