Particle-laden flow refers to fluid flow in which particles are suspended and transported by the fluid. The behavior of particle-laden flows can be complex and depends on various factors such as the size, shape, and density of particles, fluid properties, and flow conditions. Understanding the different phases of particle-laden flows is important for predicting and controlling the behavior of such flows in various industrial and environmental applications. Particle-laden flows are often classified by concentration, size, and suspension mechanisms with various multi-fidelity and multi-physics models available to model these various configurations. The choice of model depends on the dominant physics and its computational tractability. At the fluid macroscale, fluid-solid interactions are treated in a fully or partially resolved manner. The fluid influences the solid phase dynamics while the solid phase is used to update the fluid domain. At the mesoscale, the discrete resolution of solid particle contact is still tractable, but the particle length scale is comparable to the scale of the fluid elements. At this level, particle effects are introduced by modifying the equations of motion of the continuum phase while using the discrete solid particles to resolve a porosity field. The microscale considers grain sizes small enough that the solid phase must be treated from a continuum perspective. At this scale, individual solid particles cannot tractably be resolved and the solid phase dynamics are captured by the evolution of a concentration field. Modifying material properties based on the solid concentration is the primary coupling mechanism at this scale. This work presents a three-tier multiphase model that incorporates all three scales in a macro-meso-microscale GPU framework, thus enabling the unified treatment of particle-laden flows. This allows for the development of complex coupling models that allows for interactions between the three tiers while also allowing the solid mass to move between phases.