Abstract: The performance of high-precision orbit, clock, and Observable Specific Bias (OSB) products released by different analysis centers of the International GNSS Service (IGS) is evaluated by multi-frequency and multi-constellation Precise Point Positioning Ambiguity Resolution (PPPAR). The products from various IGS analysis centers, including the European Space Agency (ESA) as a reference, Wuhan University (WUM), GeoForschungsZentrum (GFZ), Center for Orbit Determination in Europe (CODE), and Centre National d'Etudes Spatiales/Groupe de Recherche de Géodésie Spatiale (CNES/GRG), is compared. Convergence time, time to first fix (TTFF), and positioning accuracy of various PPP-AR configurations in various directions are analyzed. The results show that orbit and clock of WUM demonstrated the best performance with setting ESA products as a reference. For dual-frequency dual-system PPP-AR, WUM, GFZ, and CODE showed comparable performance, with convergence times ranging from 1.7 to 2.2 minutes and TTFFs of approximately 3.1 to 3.3 minutes. In terms of positioning accuracy, WUM, GFZ, and CODE achieved accuracies of approximately 1.2 to 1.3 cm in the east direction, 1.1 to 1.6 cm in the north direction, and 2.6 to 2.8 cm in the up direction. Due to differences in the solution station network and model consistency, the PPP-AR based on CNES IGS products and PPP-Wizard project products exhibited poorer performance in terms of positioning accuracy and convergence time. For dualfrequency triple-system PPP-AR, WUM and GFZ products had similar convergence times of about 0.8 to 1.0 minutes and TTFFs of approximately 1.6 to 1.7 minutes, while CNES had longer convergence times of 2.0 to 2.3 minutes and a TTFF of 3.5 minutes. WUM achieved the best convergence accuracy, with accuracies of approximately 1.1 cm in the east direction, 1.0 cm in the north direction, and 2.3 cm in the up direction. In multi-frequency positioning, compared to dualfrequency dual-system PPP-AR using Galileo and BDS-3, the triple-frequency solution showed accelerated convergence speeds by 32.6%, 35.1%, and 34.1% in the east, north, and up directions, respectively, and a faster TTFF by 22.8%. However, as the number of frequency points increased further, the improvement margin gradually decreased. Quad-frequency PPP-AR showed faster convergence speeds of 24.1%, 20.8%, and 22.2% in the east, north, and up directions compared to triple-frequency PPP-AR, and a faster TTFF by 13.6%. There is no significant further improvement observed with five-frequency PPP-AR. CNES products generally had slower convergence speeds than WUM and did not support five-frequency PPP-AR solutions for Galileo. The study highlights the varying performance of high-precision products from different IGS analysis centers in PPP-AR. WUM demonstrated superior performance in terms of convergence time, TTFF, and positioning accuracy. These findings provide insights into the strengths and limitations of current high-precision GNSS products and inform future developments in PPP-AR techniques.