Spam Bayes is an open-source project to create the ultimate spam filter. Initially based on the work of Paul Graham, the central algorithm employs a Bayesian filter to classify and sort messages. Initially, the filter must be 'trained' by the user, who classifies mail as spam or non-spam. After a sufficient amount of training, the algorithm runs on its own with an extremely high success rate.

The engine has been built in to plug-ins for Outlook, Eudora, and other mail applications, and a variety of commercial vendors offer free, personal-use spam filters that incorporate Spam Bayes technology.