Abstract Optimizing the overall NADPH turnover is one of the key challenges in various value-added biochemical syntheses. In this work, we first analyzed the NADPH regeneration potentials of common cell factories, including Escherichia coli, Saccharomyces cerevisiae, Bacillus subtilis, and Pichia pastoris across multiple environmental conditions and determined E. coli and glycerol as the best microbial chassis and most suitable carbon source, respectively. In addition, we identified optimal cofactor specificity engineering (CSE) enzyme targets, whose cofactors when switched from NAD(H) to NADP(H) improve the overall NADP(H) turnover. Among several enzyme targets, glyceraldehyde-3-phosphate dehydrogenase was recognized as a global candidate since its CSE improved the NADP(H) regeneration under most of the conditions examined. Finally, by analyzing the protein structures of all CSE enzyme targets via homology modeling, we established that the replacement of conserved glutamate or aspartate with serine in the loop region could change the cofactor dependence from NAD(H) to NADP(H).