This paper is concerned with the optimality and duality for a class of nonconvex fuzzy optimization problems under granular differentiability. We first derive optimality conditions for fuzzy optimization problems with fuzzy inequality constraints based on granular differentiability and granular F-convexity. Then, Wolfe type and Mond-Weir type dual models corresponding to the primal problems are described, where their weak, strong, and strict converse dual theorems are explored to illustrate the connection between the optimal solutions of the primal and the dual models. Meanwhile, several examples are provided to support the corresponding theoretical results.