The two chronicles were produced within eight years of each other for different audiences. The Kojiki used Chinese characters bent to record Japanese sounds and syntax, which made it difficult even for Japanese readers. The Nihon Shoki was written in classical Chinese that a scholar in Chang'an or in the Korean kingdoms could read, and it was written to be read that way. Japan in the early eighth century was a new polity, freshly equipped with a legal code and a Chinese-style capital, making a case for its standing in a Chinese world order. A dynastic history in the correct language was part of that case.
The compilers drew on a wide range of material and, unusually, they named some of it. Korean sources are cited by title, including the Paekche annals and other Paekche records now lost except as quoted here, along with Chinese histories and older Japanese compilations, court and clan genealogies, and provincial gazetteers. Prince Toneri, a son of Emperor Temmu, presented the finished work in 720; the project itself is usually traced back to Temmu's reign in the 680s.
The Chinese is not uniform. Analysis of the language across the thirty books shows different hands, some writing correct continental Chinese and some writing a Japanised version, which is one of the main tools scholars use to sort out how the work was assembled and in what order. Whole phrases are lifted from Chinese literary and historical models and applied to Japanese events, which means a striking sentence in the chronicle may be a borrowed formula rather than a record of anything.
The relationship between the two chronicles is not that of draft and revision. They share material but neither is derived from the other, and they diverge on names, sequence and emphasis often enough that both are assumed to draw on a common pool of earlier written and oral sources, now lost.