The black-box problem of artificial intelligence （AI） is troubling users to accept AI for knowledge creation. The explainable AI is one of the important solutions to solve the problem. However，existing literature has rarely explored how the explainability of AI affects users' acceptance of AI for knowledge creation. Therefore，this study focused on exploring the question，including the path mechanism of explainability affecting users' acceptance of AI for knowledge creation，and the moderating effect of user characteristics on the path. This paper proposed some theoretical hypotheses and conducted the structural equation modeling and hierarchical regression analysis on 425 questionnaire data to test the hypotheses. The results showed that the three dimensions of explainability，i.e.，completeness，format，and currency，have an influence on users' acceptance of AI for knowledge creation；the influence of explainability on users' acceptance of AI for knowledge creation is indirect，with perceived usefulness and perceived ease of use playing a mediating role. The results also showed that the influence of explainability on users' acceptance of AI for knowledge creation is moderated by user characteristics such as education level，usage experience，and position. This study will not only contribute to the theories of AI knowledge creation and AI explainability theory by providing a user acceptance model based on the explainability，but also provide insights for enterprises to correctly play the role of AI explainability and promote AI knowledge creation.