• δ-to-γ phase transformation model for laser powder bed fusion was proposed. • Lower Cr eq /Ni eq and higher laser power accelerate δ-to-γ transformation. • Heat affected zone shows more phase transformation due to reheating by laser scans. • The model predicts printed microstructure with high accuracy (R 2 = 0.9405). In laser powder bed fusion (LPBF), complex and spatially varying thermal histories emerge due to repetitive laser scanning and localized heat accumulation, which complicate the prediction of microstructure evolution. These thermal histories result in heterogeneous microstructures composed of δ-ferrite, α’-martensite, and retained austenite in martensitic stainless steels fabricated by the LPBF. In particular, the volume fractions of α’-martensite and δ-ferrite are influenced by the δ-ferrite to γ-austenite transformation, which is affected by powder chemistry and local cooling rates. This study proposes a modeling framework to predict the δ-ferrite to γ-austenite transformation in a representative alloy. The numerical model integrates thermodynamic data and thermal simulations to perform time-dependent thermodynamic calculations based on classical nucleation and growth theories along simulated thermal histories. It successfully captures the phase transformation under ultrafast cooling conditions, showing strong agreement with experimental observations (R 2 = 0.9405). These findings offer a novel framework for understanding and controlling phase evolution in additively manufactured martensitic stainless steels, enabling site-specific microstructure control to optimize mechanical performance.