FastME 2.0 is a comprehensive, accurate, and fast distance-based phylogeny inference program. It is based on balanced minimum evolution (BME), the principle of Neighbor Joining (NJ). FastME improves upon NJ by using efficient topological moves, including Nearest Neighbor Interchange (NNI) and Subtree Pruning and Regrafting (SPR). The new version includes SPR, which allows for more powerful tree improvements than NNI. FastME offers various features such as distance estimation for DNA and proteins with multiple models, bootstrapping, and parallel computing. It is available through command-line, PHYLIP-like, and a web server interface. Distance-based algorithms infer phylogenies from pairwise distance matrices and are known for their speed and accuracy. They account for probabilistic substitution models, making them useful despite being less accurate than likelihood-based methods. NJ is a greedy algorithm that builds trees through iterative agglomeration. FastME 1.0 introduced fast BME-based algorithms to construct initial trees and improve them using NNI. These algorithms were shown to significantly improve accuracy compared to NJ. FastME 2.0 includes SPR, which allows for more efficient tree search by considering O(n²) alternative trees. The computational cost of SPR-based tree search is O(kn²), similar to NJ's O(n³). Real data experiments show that SPR improves accuracy compared to NJ and NJ+NNI. FastME 2.0 also includes various tree-building algorithms that optimize BME and OLSME, and provides distance estimation for DNA and protein sequences using different models. It supports bootstrapping and analysis of multiple datasets in a single run, and offers parallel computing via OpenMP. FastME 2.0 includes a PHYLIP-like interface, command-line interface, and a web server. It is an open-source C program available for multiple operating systems. FastME 2.0 is a comprehensive tool for phylogeny inference using distance-based methods.FastME 2.0 is a comprehensive, accurate, and fast distance-based phylogeny inference program. It is based on balanced minimum evolution (BME), the principle of Neighbor Joining (NJ). FastME improves upon NJ by using efficient topological moves, including Nearest Neighbor Interchange (NNI) and Subtree Pruning and Regrafting (SPR). The new version includes SPR, which allows for more powerful tree improvements than NNI. FastME offers various features such as distance estimation for DNA and proteins with multiple models, bootstrapping, and parallel computing. It is available through command-line, PHYLIP-like, and a web server interface. Distance-based algorithms infer phylogenies from pairwise distance matrices and are known for their speed and accuracy. They account for probabilistic substitution models, making them useful despite being less accurate than likelihood-based methods. NJ is a greedy algorithm that builds trees through iterative agglomeration. FastME 1.0 introduced fast BME-based algorithms to construct initial trees and improve them using NNI. These algorithms were shown to significantly improve accuracy compared to NJ. FastME 2.0 includes SPR, which allows for more efficient tree search by considering O(n²) alternative trees. The computational cost of SPR-based tree search is O(kn²), similar to NJ's O(n³). Real data experiments show that SPR improves accuracy compared to NJ and NJ+NNI. FastME 2.0 also includes various tree-building algorithms that optimize BME and OLSME, and provides distance estimation for DNA and protein sequences using different models. It supports bootstrapping and analysis of multiple datasets in a single run, and offers parallel computing via OpenMP. FastME 2.0 includes a PHYLIP-like interface, command-line interface, and a web server. It is an open-source C program available for multiple operating systems. FastME 2.0 is a comprehensive tool for phylogeny inference using distance-based methods.