Paleomagnetic data enables the global reconstruction of the geomagnetic field, allowing the investigation of significant events like polarity reversals and excursions. When compared to prior polarity reversals, the most recent one, the Matuyama-Brunhes (MB), is the best recorded reversal in terms of number of available paleomagnetic data. Nevertheless, several of these data have poor age control, and they are not distributed equally worldwide. Few global models have been presented for the MB; the most recent is the GGFMB (Global Geomagnetic Field Model for the MB reversal). Limitations imposed by input data and subjective assumptions about the data that are made in modelling restrict the resolution and reliability of these models. This study presents a suite of eight additional global models that reconstruct the magnetic field during the interval 700–900 ka ago, including the MB reversal and Kamikatsura (KKT) excursion. Through model comparisons, the robustness of the models in resolving MB reversal characteristics is assessed. The majority of models indicate that the reversal was mainly driven by the axial dipole field contribution gradually decreasing, while non-dipole parts slightly increased. At the core-mantle boundary, two high-latitude reverse flux patches appear at the beginning of the reversal, and it seems like a few precursors in the form of regionally seen transitional field occurred, related to variations in the decaying dipole moment. The main global polarity change occurred close to 778 ka, with the axial dipole quickly strengthening in the opposite direction in the following, completing the full polarity transition. All the models confirm the previously reported asymmetry of slow dipole decay and fast recovery, and indicate that the dipole moment was clearly lower in the late Matuyama than the early Brunhes. The whole reversal process occurred on average between 800 and 770 ka, with a duration of approximately 30 kyr. Out of four apparent excursions discovered in some of the models between 900 and 800 ka, the KKT excursion (890–884 ka), can be confirmed as a robust magnetic field feature. Additional, well dated paleomagnetic records in particular from the southern hemisphere are required to confirm several details suggested by the models that should only be interpreted with caution so far.