External aerodynamics play a crucial role in the performance of the design in many engineering problems, e.g. wind turbines, aircraft etc. In this article we present the coupling between a Lagrangian and an Eulerian solver for efficient simulations in this field of engineering.. Lagrangian solvers are getting more and more famous in the field of external aerodynamics. Their non-dissipative nature and the available acceleration techniques are two of their main advantages. However, the isotropy of their elements still poses a barrier in resolving regions with high anisotropy, such as boundary layers. On the other hand, Eulerian solvers can efficiently deal with anisotropy, but suffer from numerical diffusion, which in most cases leads to dense meshes downstream to accurately convect the vorticity in the wake, sacrificing computational speed. Eulerian and Lagrangian solvers can be coupled in a hybrid manner creating solvers that can accurately predict the vorticity generation, while they evolve the flow downstream efficiently. Here, the coupling strategy is mainly based on the work of Palha et al. [1], but replacing the explicit Finite Element Method, with an implicit Finite Volume Solver, implemented in Open- FOAM. A domain decomposition technique [2] is used for dividing the computational domain into the Eulerian and the Lagrangian part, while the coupling between the two component solvers takes place every time-step. Velocity and pressure gradient boundary conditions for the Eulerian solver are directly calculated by the Lagrangian particles, and the Lagrangian part is evolved in two steps (convection and then diffusion) [3]. Different acceleration techniques, such as Fast Multipole Method, GPU calculations, OpenMP codes etc. have been implemented making the solver more efficient. The hybrid solver is tested on an actuator line case and compared to a pure Eulerian solver (pimpleFoam solver of OpenFOAM). The results reveal that the hybrid solver can accurately predict the velocity profiles (relative errors < 1%) while they reduce the computational time sig- nificantly. Although a comparison on the computational speed is not straightforward, a speedup between 10 and 20 times is observed in the specific case.