PlatEMO is a MATLAB platform for evolutionary multi-objective optimization, offering over 50 multi-objective evolutionary algorithms (MOEAs) and 110 multi-objective test problems, along with performance indicators. It provides a user-friendly GUI for comparing algorithms, collecting results in Excel or LaTeX, and is fully open-source, enabling users to develop new algorithms. The platform includes a comprehensive set of functions for algorithm execution, problem-solving, and performance evaluation. It supports both GUI and command-line operations, allowing users to run MOEAs on MOPs with customizable parameters. PlatEMO is designed for ease of use, extensibility, and flexibility, with modular architecture that allows for the addition of new algorithms, problems, operators, and performance indicators. The platform is open-source, with source code available online, and is intended to facilitate research and application in evolutionary multi-objective optimization. It includes a variety of performance indicators such as coverage, generational distance, hypervolume, and inverted generational distance. The platform also provides a powerful experimental module for statistical analysis and benchmarking of MOEAs. Future developments aim to enhance the platform with more modules and functionalities for preference optimization, dynamic optimization, and noisy optimization. The paper describes the architecture, usage, and extension methods of PlatEMO, highlighting its advantages in supporting research and development in multi-objective optimization.PlatEMO is a MATLAB platform for evolutionary multi-objective optimization, offering over 50 multi-objective evolutionary algorithms (MOEAs) and 110 multi-objective test problems, along with performance indicators. It provides a user-friendly GUI for comparing algorithms, collecting results in Excel or LaTeX, and is fully open-source, enabling users to develop new algorithms. The platform includes a comprehensive set of functions for algorithm execution, problem-solving, and performance evaluation. It supports both GUI and command-line operations, allowing users to run MOEAs on MOPs with customizable parameters. PlatEMO is designed for ease of use, extensibility, and flexibility, with modular architecture that allows for the addition of new algorithms, problems, operators, and performance indicators. The platform is open-source, with source code available online, and is intended to facilitate research and application in evolutionary multi-objective optimization. It includes a variety of performance indicators such as coverage, generational distance, hypervolume, and inverted generational distance. The platform also provides a powerful experimental module for statistical analysis and benchmarking of MOEAs. Future developments aim to enhance the platform with more modules and functionalities for preference optimization, dynamic optimization, and noisy optimization. The paper describes the architecture, usage, and extension methods of PlatEMO, highlighting its advantages in supporting research and development in multi-objective optimization.