Almost all current and future high-contrast imaging instruments will use a Pyramid wavefront sensor (PWFS) as primary or secondary wavefront sensor. The main issue with the PWFS is its nonlinear response to large phase aberrations, especially under strong atmospheric turbulence. In this talk, we will present closed-loop lab results of a nonlinear reconstructor for the unmodulated PWFS of MagAO-X based on Convolutional Neural Networks. We show that our nonlinear reconstructor has a dynamic range of >600 nm rms, significantly outperforming the linear reconstructor that only has a 50 nm rms dynamic range. The reconstructor behaves well in closed-loop and can obtain >80% Strehl under a large variety of conditions and reaches higher Strehl ratios than the linear reconstructor under all simulated conditions. The CNN reconstructor implementation also achieves the theoretical sensitivity limit of a pyramid wavefront sensor showing that it does not lose its sensitivity in exchange for dynamic range. The current CNN’s computational time is 690 us which enables systems to run at >1 kHz. We will also discuss the real-time implementation on MagAO-X and show preliminary on-sky tests.