To study electrochemical lithiation and delithiation kinetics in silicon-graphite composite anode for lithium-ion batteries, a discrete diffusion particle model (DDPM) is developed and validated. Our recent study indicates conventional single-particle model shows limited representation of mixed chemistry. This model solves solid-phase lithium diffusion dynamics and transfer kinetics in a discrete diffusion particle system, which can be used to address the effects from composition of active particles of different chemistry, particulate morphology and size distribution. Half cells are built and tested for modeling parameter identification and validation. Tests include constant-current / dynamic charge and discharge at various temperatures. Results show well captured characteristics of inhomogeneous use of different materials, especially at state of charge region close to 0% and 100%. Comparison between modeling and testing suggests the model gives a reasonable accuracy and could be used to quantify the impacts of electrode design parameters, providing insights for improving the performance of the mixture negative electrodes.