The generalized answer is this.
The energy consumed do perform any task can be broken down into two categories: the energy it takes to do the REAL work, and the energy that is wasted in doing it. So with that in mind, know that you CANNOT SAVE ANY ENERGY THAT WAS NOT BEING WASTED.
So, the energy savings that is achievable using a VFD comes from DECREASING the energy that is wasted in VARIABLE flow processes, and ONLY those that use what are called "centrifugal" machines, otherwise known as "quadratic". The quadratic term comes from the fact that in that type of machine, energy used is related to a quadratic equation to to speed of the motor running it, being the cube of the speed reduction. So for example on a centrifugal pump, if I run the pump at 50% speed, the energy NEEDED by the pump decreases to .5 x .5 x .5 = .125, or at 1/2 speed the load requires 1/8 power. Applied to that variable flow application then, if I am using SOME OTHER MEANS to vary the flow from this pump, that other means will have losses associated with it. In pumps where you use valves for example, the valve causes a pressure drop across it, which represents an energy loss in heat in the valve and pipes due to friction and turbulence (which you may never notice because it is carried away with the fluid). Energy use does still reduce by reducing the FLOW, but at a more direct ratio, closed to 1:1 (meaning at 1/2 flow it takes 1/2 power). So by removing the valve and changing the motor speed, you remove those losses in the valve and take advantage of the DECREASED energy input for the same flow rate.
Overall, this is a concept in physics referred to as pne of the "Afinity Laws", which you can look up and research further. A similar principal exists for centrifugal fans as well, SOME types of centrifugal compressors and, perhaps in your case, a CENTRIFUGAL mixer. Whether or not your mixer can be considered a centrifugal load is entirely up to the mixer design. Paddle mixers that are essentially a pump impeller without a volute, are often considered centrifugal. Axial propeller type mixers are often not. You should be able to get your mixer supplier to answer that for you, they will (should) know.
But here is the caveat to this: if you are not VARYING the speed of the mixer, meaning that SOMETIMES you need a higher speed and MOST of the time you run at a lower speed, then you will actually NOT realize any energy savings, compared to just using a CORRECTLY sized motor for the load. In other words if that 185kW motor was sized for a task that you NEVER perform in your process, and you are only using it at 100kW, replacing it with a 100kW motor will save you MORE energy than using a VFD on the larger motor.
On all other types of machines that are NOT centrifugal, there is NO appreciable* savings that happen by magic just by applying a VFD to the motor, despite what unscrupulous salespeople might try to convince you of. You cannot violate the laws of physics. As I said in the beginning, motors only use the energy required by the load, plus any losses. Just installing a VFD does not remove losses in the motor itself, so if you cannot remove other load related losses, you will not save energy.
* I say "appreciable" here because there are a few "tricks of the trade" used by VFDs and other power electronic devices to reduce a SMALL portion of the magnetic losses inside of the motor, but only IF it is less than 50% loaded for more than 50% of the time. Those savings are very minuscule (a fraction of a fraction of a fraction) and often grossly over stated in marketing literature and "testimonials" by ignorant users, but there are also losses INTRODUCED by the power electronics tht are decidedly ignored in those statements. Don't bother considering those over stated savings, they are not worth the expense of the VFD or other device. In a VFD, those can be used to further ENHANCE the affinity law savings, but are not worth it stand-alone.