Fpocket is an open-source platform for detecting ligand pockets in proteins, based on Voronoi tessellation and alpha spheres. It is written in C and is designed for efficient and scalable pocket detection, scoring, and descriptor extraction. The software includes three main programs: Fpocket for pocket identification, Tpocket for benchmarking pocket detection, and Dpocket for extracting pocket descriptors. Fpocket uses alpha spheres to identify potential binding sites by analyzing the geometry of protein surfaces. It is particularly effective in detecting pockets that are likely to bind small molecules, and it outperforms existing methods in terms of speed and accuracy. The software is freely available under the GNU GPL license and is suitable for further development and research in protein pocket identification and drug discovery. Fpocket provides a modular framework for large-scale pocket detection, descriptor extraction, and benchmarking, making it a valuable tool for the scientific community. The software has been tested on various datasets and has shown high performance in identifying pockets, particularly in the top-ranked pockets. It is also efficient in terms of computational performance, making it suitable for high-throughput screening and the construction of cavity databases. Fpocket is an open-source platform that offers a robust and flexible solution for the identification of protein pockets, which is essential for drug discovery and molecular modeling.Fpocket is an open-source platform for detecting ligand pockets in proteins, based on Voronoi tessellation and alpha spheres. It is written in C and is designed for efficient and scalable pocket detection, scoring, and descriptor extraction. The software includes three main programs: Fpocket for pocket identification, Tpocket for benchmarking pocket detection, and Dpocket for extracting pocket descriptors. Fpocket uses alpha spheres to identify potential binding sites by analyzing the geometry of protein surfaces. It is particularly effective in detecting pockets that are likely to bind small molecules, and it outperforms existing methods in terms of speed and accuracy. The software is freely available under the GNU GPL license and is suitable for further development and research in protein pocket identification and drug discovery. Fpocket provides a modular framework for large-scale pocket detection, descriptor extraction, and benchmarking, making it a valuable tool for the scientific community. The software has been tested on various datasets and has shown high performance in identifying pockets, particularly in the top-ranked pockets. It is also efficient in terms of computational performance, making it suitable for high-throughput screening and the construction of cavity databases. Fpocket is an open-source platform that offers a robust and flexible solution for the identification of protein pockets, which is essential for drug discovery and molecular modeling.