<p>The continuous miniaturization of engineering devices demands predictive models for thermoelastic wave dispersion in microscale composite beams that accommodate both size dependence and finite-speed heat conduction. Yet the wave dispersion of functionally graded (FG) Timoshenko microbeams co-reinforced with graphene platelets (GPLs) and carbon nanotubes (CNTs) has rarely been examined under the concurrent action of nonlocal strain gradient (NSG) elasticity and Lord–Shulman (L-S) generalized thermoelasticity. In this work, we establish a coupled NSG–L-S framework for FG GPLs/CNTs-reinforced Timoshenko microbeams. The grading is represented by a layer-wise through-thickness distribution with three patterns (FG-A, FG-O, and FG-X), and effective properties are evaluated using the Halpin-Tsai model combined with the rule of mixture. A harmonic wave assumption reduces the governing equations to a homogeneous algebraic system, and the dispersion relations of coupled thermoelastic bending- and shear-dominated wave modes are obtained from the nontriviality condition. Model validity is verified by reducing the formulation to benchmark cases and matching published dispersion solutions. Results show that FG-A produces the largest increase in dispersion frequencies; for the same filler content, GPLs enhance stiffness more effectively than CNTs. The size parameters exhibit competing roles, with nonlocality causing softening and the strain gradient effect producing stiffening. Thermal relaxation further lowers dispersion frequencies, especially at higher wavenumbers. Overall, the proposed approach provides a reusable theoretical framework for elucidating how reinforcement grading, size effects, and thermal relaxation jointly influence thermoelastic wave dispersion in FG microbeams.</p>