Wind turbine rotor optimizing has been the topic of significant amount research since the start of the wind turbine industry. For many years the target has been to make the most power efficient rotor (max CP), but that is changing to rotor optimization where loads constraints are being taken into account. We have previously made a relatively simple analytical model which relates the rotor loading and the rotor power at every radial position. The model can be used to study the trade-off between loading and power and the optimal load distribution can be found for a given load constraints (e.g. thrust, flap moment, etc.). But the model comes with assumption/approximation. In order to capture more of the physics a surrogate model will be build on top of the analytical model. The surrogate model is computationally fast to evaluate and it works as a correction model for the analytical model. The surrogate model will be trained with two different CFD model, which differs in computational speed and fidelity - both of them being 3D CFD model. The first CFD model is an Actuator Disk (AD) model which only requires to set the forces at the rotor disk, which is opposed to specifying a blade geometry. The second CFD model is a rotor simulation where the blade geometry is resolved. The AD model has a shorter computational time but it does not capture as much of the physics. The work will focus on how to incorporate the multi fidelity information in a single design framework. Using the multi fidelity approach the optimization strategy used for the analytical model can still be used which made the rotor optimization problem simpler and it can reduce the amount of optimization parameters. The amount of optimization parameters is a common problem in optimization especially when the amount of simulations that can realistically be made is on the order of 50-100 and the number of parameters is more than 10 (the problem is known as The curse of dimensionality).