Abstract The problem of long convergence time restricts the development of precise point positioning (PPP)technology in the application field. In order to improve the positioning accuracy and convergence speed of PPP, we propose an ionosphere-constraint model based on uncombined and unconstrained model, and the high-precision regional ionospheric a priori information constraint method. We use the observation data of 183 reference stations in the west coast CORS network of the United States to model the regional ionosphere, obtain high-precision ionospheric delay information, and compare and analyze through experiments, the PPP positioning performance of the three algorithms: ionosphere-free, uncombined and unconstrained and ionosphere-constraint model. The experimental results show that the additional ionospheric constraint algorithm has better model geometric strength. Compared with the first two models, the PPP 10 minutes before the positioning process, this algorithm can significantly improve the convergence speed and positioning accuracy of PPP positioning.