I'm already using BEM to get Cp, but when i use CFX to get similar values it don't happen.s

Sounds pretty reasonable. A BEM software works very different than CFD. BEM under predicts the overall performance of a wind turbine and over predicts the peak power. Also CFD is as accurate as the model and boundary conditions used.
I am also modelling wind turbines and in my opinion the only safe way to judge the BEM or CFD results is an experimental reference.

Even compared with experimental reference, my Cp in
CFD is so different.
Basically, at 5RPM i get Cp of 0.13 in CFX and 0.17 in BEM. But, at 10RPM i get Cp of 0.61 in CFX and 0.42 in BEM. So, is evident something is wrong with my simulation in CFX.

Are you sure about the RPM you use? What's your reference CP(@what RPM, wind speed)?

I'm sure about RPM i'm using. My reference of Cp is BEM. Im getting Cp of 0.65 at 10RPM, 2.5m/s of wind speed and my turbine has a diameter of 10m. So, in BEM i get a CP of 0.4 to theses conditions. So, something is wrong in my simulation.

I guess your simulation has some problems. You set wind speed at 2.5m/s, if your Tip speed ratio is 7, diameter is 10m. The rotating speed is 33.44RPM, not 10RPM.

In my simulation, from cut-in speed (2.5m/s) to around 10m/s, the Cp is almost same as in BEM. But, in high wind speed such as above 12m/s, the Cp is very high. The reason is I didn't stall the rotating speed in a fix number (such as 60 RPM). I just calculate RPM according to Tip Speed Ratio and wind speed.

How is your domain? Big enough? Moreover, meshing is another issue. You'd better refine your model, if it is coarse.

I'm doing a curve of Cp X Tip speed ratio, in this case i'm setting only to 2.5m/s of wind speed and increasing the rotation to see how Cp increase.
My domain is big enough as shown in CFX's lectures and my mesh is good aswell. I changed the distance between the blades and interface, and cp decrease. So, i'm gonna try this way now. Do you know how far your blade was from the interface between rotationary and stationary domain?

Your blade's diameter is 10meter. At 2.5m/s, 5RPM means tip speed ratio(TSP) is about 1, Cp is 0.13. But, in 10RPM, TSP is about 2, Cp reaches to 0.42 in BEM. I think 0.42 is very high if TSR is 2. You know Betz limit is 0.59. Generally,the optimal TSR is 6 to 7.

I created rotational and stationary domain in AWB. I know it.

By the way, if possible, you can post your simulation model. I think more guys can help you.:)