<p>This article develops an adaptive prescribed performance tracking control scheme for nonlinear multi-agent systems (MASs) with asymmetric dead-zone output. To this end, a new adaptive prescribed performance controller is formulated using an novel global prescribed time function (GPTF) and error-based non-singular scalar transformation. Compared to the existing results on prescribed performance control (PPC) or global prescribed performance control (GPPC) for MASs, the presented scheme can guarantee that the prescribed time and prescribed accuracy are directly preassigned by designers for any initial values. Additionally, an improved smooth approximation model for the asymmetric dead-zone is devised, thereby effectively improving the approximation performance. Furthermore, to reduce the computational load, the command filter has been designed, and the compensation signals are crafted to counteract the impact of boundary errors generated by the command filter despite the existence of unknown virtual control coefficients. The simulations illustrate the feasibility of the constructed scheme.</p>