Objectives of this research are to identify types and levels of code-mixing and the most dominant one in Deddy Corbuzier’s Podcast. The researcher employs descriptive qualitative method by downloaded podcast video from youtube and write down the script and classified the types of code-mixing and levels of code-mixing in utterances of the dialog between Deddy and Azka in podcast channel video. After collecting and selecting the data, the researcher starts to analyze the data by using Muskyen and Suwito's qualitative data analysis method. The method consists of three steps, namely data reduction, data display, and verification and conclusion drawing. The results show that in Deddy Corbuzier’s Podcast Channel there were 77 utterances of Code-Mixing. They are 77 data of types of code-mixing and 90 data of levels of code-mixing The dominant type of code mixing is “Altetnation”. While the dominant levels of code-mixing is “Clause”. The researcher found that there are types of code mixing, they are 23 data of insertion, 33 data of alternation, and 21 data of congruent Lexicalization. Then, The reseacher found that levels of code-mixing, there are 30 data of word, 20 data of phrase, 1 data of repitition, 1 data of baster, 38 data of clause, and no idioms levels Based on the result above, it was concluded that all types and levels of code-mixing that was found in Deddy Corbuzier’s Podcast Channel.