Abstract Among the factors influencing the animal gastrointestinal tract microbiome (AGM) diversity, diet and phylogeny have been extensively studied. However, what made the studies particularly challenging is that diet characteristics per se are product of evolution, and hence totally disentangling both effects is unrealistic, likely explaining the lack of consensus in existing literatures. To further explore microbial diversity and host-phylogeny–diet relationships, we performed a big-data meta-analysis with 4903 16S rRNA AGM samples from 318 animal species covering all six vertebrate and four major invertebrate classes. We discovered that the relationship between AGM-diversity and phylogenetic timeline (PT) follows a power-law or log-linear model, including diet specific power-law relationships. The log-linear nature predicts a generally rising trend of AGM diversity along the evolutionary tree starting from the root, which explains the observation why Mammalia exhibited the highest AGM-diversity. The power-law property suggests that a handful of taxa carry disproportionally large weights to the evolution of diversity patterns than the majority of taxa, which explains why the species richness of Insecta was significant different than those from the other nine classes. Finally, we hypothesize that the diversity—PT power-law relationship explains why species-abundance distributions generally follow highly skewed probability distributions.