cfSNV is an ultra-sensitive and accurate somatic SNV caller designed for cfDNA sequencing. Taking advantage of modern statistical models and machine learning approaches, cfSNV provides hierarchical mutation profiling and multi-layer error suppression, including error suppression in read mates, site-level error filtration and read-level error filtration. The R package of cfSNV is available on GitHub. In addition, we also implemented a Docker image of the cfSNV package, which is designed such that researchers and clinicians with a limited computational background can easily carry out analyses on both high-performance computing platforms and local computers. The docker image is available on GitHub as well. Both R package and Docker image of cfSNV can be freely used for educational and research purposes by non-profit institutions and U.S. government agencies only under the UCLA Academic Software License.