Abstract: Addressing the critical requirements of RF power amplifier systems in large-scale particle accelerators and nuclear fusion devices, this paper investigates the precise modeling and multi-objective optimization design of high-power vacuum tetrode amplifiers. A nonlinear equivalent model incorporating space current distribution, inter-electrode capacitance effects, and thermally-induced transconductance decay was established to accurately describe the electrical characteristics of tetrodes under high-frequency and high-power conditions. Static operating points were determined based on load-line analysis and the thirteen-point method. A multi-objective genetic algorithm (NSGA-II) was employed to dynamically optimize the anode voltage utilization coefficient, grid voltage, and load impedance, achieving Pareto-optimal trade-offs among efficiency, linearity (IMD3), and power consumption. Experimental validation demonstrated that the optimization achieved 150 kW output at 44.5 MHz with an efficiency of 71.35% and an IMD3 improved to −29.1 dB, representing a significant enhancement over conventional designs. This research provides a comprehensive solution from precise modeling and static design to dynamic optimization for high-power RF systems, significantly improving the performance and reliability of power amplifiers in large-scale scientific installations.