Magnetic field measurements play a crucial role in deep space exploration, contributing significantly to our understanding of planetary habitability and the space plasma environment. Among the various instruments employed in space exploration missions, the fluxgate magnetometer stands out as a widely used tool. However, its zero offset undergoes gradual changes, necessitating regular in-flight calibration. This article comprehensively reviews in-flight calibration methods for spaceborne magnetometers in deep space exploration, leveraging physical phenomena inherent to the space environment. The methods for calculating the zero offset can be divided into two categories. The first group employs formulas, including the Belcher method, Hedgecock method, Davis-Smith method, and both one-dimensional and three-dimensional mirror mode methods. Notably, the Davis-Smith method emerges as the optimal choice among these approaches. The second group employs probability-based solutions, constituting the in-flight calibration technology of the new generation, which encompasses six algorithms. This innovative technology utilizes phenomena such as Alfvén waves, magnetic mirror structures, and current sheets for in-flight calibration. Moreover, this technology can leverage interplanetary magnetic field data, specifically requiring a sufficient data amount (e.g., 4 hours) to calculate the zero offset. Compared to the Davis-Smith method, the in-flight calibration technology of the new generation demonstrates good compatibility, allowing for the simultaneous use of different physical phenomena, and scalability. This technology extends its applicability beyond solar wind, enabling in-flight calibration in the Martian magnetosheath. Currently, this technology has successfully provided calibration services for the magnetic field data of the Tianwen-1 orbiter.