Companies increasingly use Application Programming Interfaces (APIs) to create new value propositions in the digital world. However, there is little knowledge about the type of value propositions that public APIs enable companies to provide. Thus, this study analyzes data from 13,829 API descriptions published in programmableweb.com to: (i) identify and categorize current and emerging API customer value propositions, (ii) analyze the evolution of API customer value propositions, and (iii) showcase Topic Modeling as a feasible technique to study API customer value propositions. This research contributes towards API service systems research by: (i) identifying six categories of API customer value propositions, and (ii) revealing a long tail distribution of API customer value propositions that is flattening.