In this talk we discuss some new Diophantine applications of modularity results. In the first part, we discuss a refinement of the Arakelov-Faltings-Parshin method for moduli schemes of elliptic curves. We also provide some motivation. In particular, we work out explicitly the method for certain moduli schemes to improve the actual best height bounds for Mordell equations. In the second part, we discuss an e ffective Shafarevich conjecture for abelian varieties of (product) GL2-type.