Continuous Normalizing Flows (CNFs) are a family of continuous-time deep generative models. Flow Matching (FM) has been proposed for simulation-free training of CNFs based on conditional probability paths between a source and target distribution. This enables training based upon a simple regressions objective similar to the one used in diffusion models. However, in contrast to diffusion models this formulation does not require the source distribution to be Gaussian noise which makes FM naturally applicable to domain translation. Despite initial promising results on generative modelling tasks, FM research is still in its early developmental phase. In this project, the goal is to explore the application of FM to unpaired domain translation tasks including single-cell and image translation. One promising approach is to leverage the framework of optimal transport to reduce the length of learned paths in FM.