Several physical mechanisms of earthquake nucleation, such as pre-slip, cascade triggering, aseismic slip, and fluid-driven models, have been proposed. However, it is still not clear which model plays the most important role in driving foreshocks and mainshock nucleation for given cases. In this study, we focus on the relationship between an intensive earthquake swarm that started beneath the Noto Peninsula in Central Japan since November 2020 and the nucleation of the 2024 M7.6 Noto Hanto earthquake. We relocate earthquakes listed in the standard Japan Meteorological Agency (JMA) catalog since 2018 with the double-different relocation method. Relocated seismicity revealed that the 2024 M7.6 mainshock likely ruptured a thrust fault above a parallel fault where the M6.5 Suzu earthquake occurred in May 2023. We find possible along-strike and along-dip expansion of seismicity in the first few months at the beginning of the swarm sequence, while no obvious migration pattern in the last few days before the M7.6 mainshock was observed. Several smaller events occurred in between the M5.5 and M4.6 foreshocks that occurred about 4 and 2 minutes before the M7.6 mainshock. The Coulomb stress changes from the M5.5 foreshock were negative at the hypocenter of the M7.6 mainshock, which is inconsistent with a simple cascade triggering model. Moreover, an M5.9 foreshock was identified in the JMA catalog 14 s before the mainshock. Results from back-projection of high-frequency teleseismic P waves show a prolonged initial rupture process near the mainshock hypocenter lasting for ∼25 seconds, before propagating bi-laterally outward. Our results suggest a complex evolution process linking the earthquake swarm to nucleation of the M7.6 mainshock at a region of complex structures associated with the bend of a mapped large-scale reverse fault. A combination of fluid migration, aseismic slip and elastic stress triggering likely work in concert to drive both the prolonged earthquake swarm and the nucleation of the M7.6 mainshock.