Gumbel distribution functions (De Haan L. (2007) <doi:10.1007/0-387-34471-3>) implemented with the techniques of automatic differentiation (Griewank A. (2008) <isbn:978-0-89871-659-7>). With this tool, a user should be able to quickly model extreme events for which the Gumbel distribution is the domain of attraction. The package makes available the density function, the distribution function the quantile function and a random generating function. In addition, it supports gradient functions. The package combines 'Adept' (C++ templated automatic differentiation) (Hogan R. (2017) <doi:10.5281/zenodo.1004730>) and 'Eigen' (templated matrix-vector library) for fast computations of both objective functions and exact gradients. It relies on 'RcppEigen' for easy access to 'Eigen' and bindings to R.