The general model of the Naive Bayes classifier for text classification can be written as: $\hat{c}=\text{argmax}_{c\in C}\log P(c)+\sum_i \log P(w_i|c)$, where $C$ is the set of all possible classes, and $i$ is the position of word $w_i$ in the text. This is a linear model because the prediction is computed as a linear combination of input features (i.e., the words).