We present a Quick Response (QR) code style transfer network based on conditional instance regularization to retain more QR code information and incorporate multiple styles into a single model. Firstly, we introduce conditional instance regularization into the QR code style transfer network and incorporate multiple styles into a single model. It improves the efficiency of multi-style transfer training. Secondly, the weighted fusion corrected method of the styled QR code is designed to repair the damage of positioning graphics and information modules in the QR code, enhance the styled QR code’s identifiability and guarantee the integrity of information modules. Because of the shortcomings of then static QR code, a dynamic artistic style QR code is proposed. The exper- imental results show that the proposed QR code style transfer method is effective.