The model describing the forces acting on rice seeds within seed metering systems significantly impacts the design and structural parameters of rice pneumatic seed metering mechanisms. To formulate a mathematical representation of the device during seed suction, a DEM-CFD bidirectional coupling approach was utilized to simulate and model the seed-suction process. The dynamic behavior of rice seeds under the airflow at the suction port was examined, and force models for various seed-suction positions were developed. Simulation analyses were conducted under varying conditions, including different rotation speeds of the seed-suction plate and different levels of negative suction pressure. The effects of these variables on seed suction were investigated, leading to the development of a drag force model. Optimal seeding parameters and precision were determined from the simulation results. Verification and analysis were conducted through empirical experiments, and the optimal seeding performance was attained at a negative suction pressure of 1.6 kPa and a plate rotation speed of 30 r/min. The results indicated a seeding qualification rate of 95.7%, a reattachment rate of 2.4%, and a leakage suction rate of 1.9%. The simulation findings were consistent with experimental data, demonstrating that the seed-metering device satisfies the requirements for field seeding.