Abstract: The leakage inductance of high-frequency transformers is an important parameter affecting the power transmission characteristics and operational reliability of solid-state transformers, so it needs to be predicted quickly and accurately during the design stage. Due to the high computational cost of the numerical method, it is difficult to meet the requirements of quickly predicting leakage inductance during the transformer design stage. Currently, the analytical method is often used to calculate leakage inductance parameters. The calculation models based on one-dimensional magnetic field assumption ignore the influence of the winding edge effect and cannot accurately estimate leakage inductance. Although most two-dimensional calculation models consider the influence of transverse leakage flux, the solution of Poisson equation and the representation of current density in terms of series lead to an overly cumbersome calculation process. Therefore, this paper accurately extracts the leakage magnetic field in the core window based on the image principle, derives an analytical expression for the DC leakage magnetic energy in this area, and introduces frequency-dependent factors that take into account the skin effect and proximity effect of the winding to reveal the attenuation law of leakage inductance in a wide frequency range, and establishes an analytical calculation model for leakage inductance that can simultaneously consider the frequency-dependent characteristic and winding edge effect. Finally, a leakage inductance measurement platform is built and a detailed experimental study is carried out on transformer prototypes with different porosities to verify the accuracy and validity of the proposed model.