This paper describes SolveDSGE, a comprehensive Julia package designed to solve and analyze nonlinear Dynamic Stochastic General Equilibrium models. Although the package supports standard perturbation methods up to the fourth order, its originality and primary contribution resides in its implementation of various projection methods—including Chebyshev, Smolyak, hyperbolic-cross and piecewise linear approximations—to achieve global accuracy. By providing a unified interface that allows interchangeable solution initialization (homotopy) the framework significantly advances our ability to solve nonlinear models. We demonstrate the utility of the package through a series of applications, highlighting its efficiency and ease of use in modern macroeconomic research.