Here is a comparison between two Allegro hall effect-based integrated conductor current sensors, the ACS72981 for high currents (50 to 200A), and the ACS71240 for lower currents (<50A). We can see that in terms of Bandwidth, the shunt-based solutions look higher than the hall effect-based ones, but this frequency range is usually not used because having a higher frequency will induce bad accuracy due to the shift in impedance that the shunt inductance component is bringing (3nH for example for competitor B). This is explained in more detail in slide 4. This doesn’t affect the hall effect-based solution which doesn’t measure current by impedance of the conductor, but by sensing the magnetic field. The ACS72981 conductor resistance is very low in order to keep the power consumption low at high currents. If a discrete shunt solution is preferred, the shunt resistance tends to be much higher, unless the customer is willing to spend more money on a low impedance shunt resistor that would be much more voluminous too. The current range and voltage isolation is the biggest strength of the hall effect-based solution that have no comparison with the shunts. The size is a  lot more smaller as well for the 3x3mm2 QFN. It’s still pretty contained for the integrated shunt opamp, but it’s only for low power applications (<80V and <15A). And in terms of price, we would expect to have more expensive current sensors for the better performances. But the 1ku prices are very contained for the ACS71240 in QFN package that can be found at 60 cents while the other solutions can be found above $1 at the same volume with the shunt. And it’s not taking into account the overcurrent fault detection that the ACS71240 has that would require a comparator and voltage reference for the shunt-based solution. Indeed, all in all, the magnetic current sensors don’t require much work for their implementation, especially for higher voltages where isolation stages would need to be added.