Microbes play an important role in rhizosphere phosphorus (P) activation and root P absorption in low P-available soils. However, the responses of the rhizosphere microbial community to P input and its effects on P uptake by tea plants have not been widely reported. In this study, the high-throughput sequencing of the 16S rRNA gene and the ITS2 region was employed to examine the responses of tea rhizosphere microbiomes to different P input rates (low-P, P0: 0 mg·kg⁻¹ P; moderate-P, P1: 87.3 mg·kg⁻¹ P; high-P, P2: 436.5 mg·kg⁻¹ P). The results showed that the P input treatments significantly reduced the soil C: N ratio and C: P ratio compared to the P0 treatment (p < 0.05). Moreover, the P2 treatment significantly increased the soil available P, plant biomass and P content of the tea plant compared to the P0 and P1 treatments (p < 0.05). Both bacterial and fungal communities revealed the highest values of alpha diversity indices in the P1 treatment and the lowest in the P2 treatment. The dominant phyla of the bacterial community were Proteobacteria, Actinobacteria and Acidobacteria, while in the fungal community they were Ascomycota and Mortierellomycota. In addition, P input enriched the relative abundance of Actinobacteria and Proteobacteria but decreased the relative abundance of Acidobacteria. The Mantel correlation analysis showed that the fungal community was influenced by P input, whereas bacterial community was affected by the soil TC and C: N ratio. Furthermore, the P input treatments enhanced the TCA cycle, amino and nucleotide glucose metabolism, starch and sucrose metabolism, and phosphotransferase system expression, which could promote C and N cycling. On the contrary, the P input treatments negatively affected the growth of arbuscular mycorrhizal fungi. The PLS-PM model revealed that the rhizosphere bacterial and fungal communities, respectively, negatively and positively affected the P content by affecting the biomass. Meanwhile, rhizosphere microbial function profiles affected the P content of tea plants directly and positively. In summary, moderate P input favors the rhizosphere microbial diversity and functions in the short-term pot experiment. Therefore, we suggest that moderate P input should be recommended in practical tea production, and a further field test is required.