BACKGROUND: The combination of Astragalus membranaceus and Salvia miltiorrhiza (AS) is an effective prescription for treating diabetic kidney disease (DKD) in traditional Chinese medicine. Its efficacy in treating DKD has been confirmed, but the potential regulatory mechanism has not yet been fully clarified. PURPOSE: To explore the mechanism by which AS regulates the "gut-metabolism-transcription" coexpression network under the action of the "gut-kidney axis" to ameliorate DKD. METHODS: SD rats were used to establish the DKD model by injecting STZ. After AS intervention, the structure and function of the kidney and colon were observed. We sequenced the gut microbiota utilizing 16S rDNA, identified serum differential metabolites using LC‒MS/MS, and observed renal mRNA expression by RNA seq. The "gut-metabolism-transcription" coexpression network was further constructed, and the target bacteria, target metabolites, and target genes of AS were ultimately screened and validated. RESULTS: AS improved renal pathology and functional damage and increased the abundance of Akkermansia, Akkermansia_muciniphila, Lactobacillus and Lactobacillus_murinus. Fourteen target metabolites of AS were identified, which were mainly concentrated in 19 KEGG pathways, including sphingolipid metabolism and glycerophospholipid metabolism. Sixty-three target mRNAs of AS were identified. The top 20 pathways were closely related to glycolipid metabolism, and 14 differential mRNAs were expressed in these pathways. Correlation analysis showed that Akkermansia, Akkermansia muciniphila, Lactobacillus and Lactobacillus murinus were closely associated with sphingolipid metabolism, glycerophospholipid metabolism, arachidonic acid metabolism, ascorbate and aldarate metabolism and galactose metabolism. Moreover, the target metabolites and target mRNAs of AS were also enriched in five identical pathways of sphingolipid metabolism, glycerophospholipid metabolism, arachidonic acid metabolism, ascorbate and aldarate metabolism and galactose metabolism, including 8 different metabolites, such as sphingosine, and 5 different genes, such as Kng1. The 8 metabolites had high AUC prediction values, and the validation of the 5 genes was consistent with the sequencing results. CONCLUSION: Our research showed that AS can improve DKD via the "gut-kidney axis". Akkermansia muciniphila and Lactobacillus murinus were the main driving bacteria, and five pathways related to glycolipid metabolism, especially sphingolipid metabolism and glycerophospholipid metabolism, may be important follow-up reactions and regulatory mechanisms.