Inter-frequency Differential Code Biases (DCBs) account for the most significant systematical biases when sensing the earth’s ionosphere with GNSS observations and are also important correction parameters in GNSS applications of Positioning, Navigation and Timing (PNT). With the continuous modernization of the American GPS and Russian GLONASS systems, and also the rapid developments of the European GALILEO and Chinese BDS systems, there is a strong demand for precise satellite DCB products for multiple constellations and frequencies. This study proposes a new method for the precise determination of multi-GNSS triple-frequency satellite DCBs, which can be divided into three steps. The first step is to precisely retrieve slant ionospheric delays and “additional code biases” based on a newly-established full-rank triple-frequency Precise Point Positioning (PPP) function model with raw observations. Both the slant ionospheric delays and “additional code biases” contain the DCBs need to be estimated. Then an enhanced IGGDCB method is used to estimate the DCBs between the first and second frequency bands with the PPP-derived slant ionospheric delays. At last, the previously estimated DCBs between the first and second frequency bands are substituted into the “additional code biases” and DCBs between the first and third frequency bands are estimated. Multi-GNSS slant ionospheric delays from the triple-frequency PPP method are compared with those from the traditional dual-frequency Carrier-to-Code Level (CCL) method in terms of formal precision and zero-baseline experiment, in which obvious improvements are found for the PPP method. One month of data from 60 globally-distributed Multi-GNSS Experiment (MGEX) stations are selected and totally eight types of DCBs are estimated for GPS, GLONASS, GALILEO and BDS. Multi-GNSS satellite DCBs generated with the proposed method are compared with the products from different agencies, including Center for Orbit Determination in Europe (CODE), Deutsches zentrum für Luft-und Raumfahrt (DLR) and Chinese Academy of Sciences (CAS). Results show that the proposed method has good performances in multi-GNSS multi-frequency satellite DCB estimation.