The collective dynamics of active biopolymers is crucial for many processes in life, such as cellular motility, intracellular transport, and division. Recent experiments revealed fascinating self-organized patterns of diverse active filaments, while an explicit parameter control strategy remains an open problem. Moreover, theoretical studies so far mostly dealt with active chains with uniform stiffness, which are inadequate in describing the more complicated class of polymers with varying stiffness along the backbone. Here, using Langevin dynamics simulations, we investigate the collective behavior of active chains with homogeneous and heterogeneous stiffness in a comparative manner. We map a detailed non-equilibrium phase diagram in activity and stiffness parameter space. A wide range of phase states, including melt, cluster, spiral, polar, and vortex, are demonstrated. The appropriate parameter combination for large-scale polar and vortex formation is identified. In addition, we find that stiffness heterogeneity can substantially modulate the phase behaviors of the system. It has an evident destructive effect on the long-ranged polar structure but benefits the stability of the vortex pattern. Intriguingly, we unravel a novel polar-vortex transition in both homogeneous and heterogeneous systems, which is closely related to the local alignment mechanism. Overall, we achieve new insights into how the interplay among activity, stiffness, and heterogeneity affects the collective dynamics of active filament systems.