Climate warming are supposed to have irreversible effects on the biodiversity and ecosystem functions of grassland. Alpine meadow is the predominant ecosystem in Qinghai-Tibet Plateau, which has attracted much attention because of its sensitivity to global changes. However, there is still no unified understanding of the impact of warming on the relationship between diversity and productivity of alpine meadows. In this study, we conducted multi-gradient warming, used fiberglass open-top chambers (OTCs) to raise the temperature, based on the investigation of community structure and productivity, we evaluated the effect of field simulated warming on the relationship of species diversity-primary productivity. Our result showed that gradient warming reduced species richness, diversity and dominance, inhibited the accumulation of above-ground biomass (AGB) and below-ground biomass (BGB), which was more significant under higher warming (P < 0.05). AGB had a significant linear positive correlation with Margalef index (R) and Shannon-Wiener index (H) (P < 0.05), but was weakly dependent on Simpson index (H′) and Pielou index(E), while BGB was not affected by species diversity during the experiment. Our results revealed that simulated warming weakened the dependence of AGB on species diversity, and aggravated the negative correlation between BGB and species diversity (P < 0.05), which was mainly caused by the decrease of R, H, H' after warming. Functional group level, we found the AGB of Forbs increased significantly with the increase of species richness (P < 0.05), but warming had little effect on the relationship between them, while the relationship between species richness and AGB of Poaceae was significantly strengthened under warming. It concludes that the interaction between Poaceae and Forbs leads to a gradual decrease in above-ground biomass at the community level with the increase of species richness under warming. Our work provides an experimental data source for BEF research and verification under climate change scenarios, and an important reference for predicting and dealing with the impact of global climate change on adaptive management and grassland protection.