The purpose of this study is to build a training model of teachers’ intelligence literacy based on DL (Deep Learning) and verify its effectiveness and feasibility through simulation experiments. In this study, DL technology is introduced into the field of teachers’ intelligence literacy training, which realizes the accurate identification of teachers’ intelligence literacy level and the recommendation of personalized training content. Through the test results of four experimental indicators, it is verified that the DL-based teacher intelligence literacy training model proposed in this paper is effective in recognition accuracy, personalized recommendation, training effect improvement and training efficiency. The experimental results show that the recognition accuracy of the model on the test set reaches 92.5%, showing high recognition accuracy; at the same time, after training, the level of teachers’ intelligence literacy has increased by 25% on average, which is significantly higher than that of untrained teachers. These results fully prove the practicability and application prospect of this method. Through this study, I hope to provide new ideas and methods for the training of college teachers’ intelligence literacy and promote the overall quality of college teachers.