A high quality factor (Q-factor) is a critical parameter determining the sensing sensitivity and frequency stability of MEMS resonators. However, given the complex energy dissipation mechanisms at the micro/nano-scale, conventional design approaches relying on geometric intuition struggle to maximize the Q-factor within an environment of multiple coupled damping sources. To address this challenge, this paper proposes a topology optimization design methodology tailored for high-Q MEMS beam resonators. First, grounded in thermoelasticity and elastic wave radiation theories, a comprehensive multiphysics simulation framework incorporating both thermoelastic damping (TED) and anchor loss is established. By integrating Perfectly Matched Layers (PMLs) with coupled thermal-structural equations, the total energy dissipation of the resonator is quantified with high precision. Subsequently, a density-based topology optimization algorithm is employed to evolve the structure of a silicon clamped-clamped beam, with the objective of maximizing the Q-factor of the fundamental mode. The study yields two novel topological configurations exhibiting significantly reduced energy loss. Physical mechanism analysis reveals that the optimized material distribution effectively interrupts transverse heat-flow pathways, thereby suppressing TED, while simultaneously inducing a “soft-clamping” redistribution of strain energy near the anchors to minimize energy leakage into the substrate. Simulation results demonstrate that, compared with a conventional solid straight beam of identical dimensions, the optimized designs achieve an approximately six-fold enhancement in the overall Q-factor. This work confirms the effectiveness of the multiphysics-coupled topology optimization strategy, providing new theoretical guidance and technical pathways for the design of high-performance MEMS resonators.