The paper is concerned with the robustness of global exponential stability of hybrid neural networks subject to noise and delay simultaneously. Given a globally exponentially stable hybrid neural network, the aim of the paper is to characterize how much delay and noise intensity hybrid neural networks can bear such that the perturbed hybrid neural network remains globally exponentially stable, in the presence of delay and noise simultaneousl...