In this paper, we describe a probabilistic swarming path planning algorithm based on optimal transport. This algorithm has been developed based on the basic concept of controlling the density of discretized regions of given environment and can be effectively applied to large-sized swarms. The proposed algorithm abstracts the given environment by using centroidal Voronoi tessellation and obtains optimal transport plan through optimal transport algorithm. Voronoi diagram based collision avoidance also considered to design a collision-free path planning algorithm. Through several simulations, we confirmed that the proposed algorithm operates successfully in various scenarios.