Following recent advancements in cryo-electron microscopy (cryo-EM) instrumentation and software algorithms, the next bottleneck in achieving high-resolution cryo-EM structures arises from sample preparation. To overcome this, we developed a graphene-based affinity cryo-EM grid, the Graffendor (GFD) grid, to target low-abundance endogenous protein complexes. To maintain grid quality and consistency within a single batch of 36 grids, we established a one-step crosslinking batch-production method using genetically modified ALFA nanobody as affinity probe (GFD-A grid). Using low concentrations of β-galactosidase-2xALFA, we demonstrated the GFD-A grid's efficiency in capturing tagged proteins and resolving its cryo-EM structure at 2.71 Å. To test its application for endogenous proteins, we engineered yeast cells with a C-terminal tandem affinity tag (3xALFA-Tev-3xFlag: ATF) at Pop6, a shared component of RNase MRP and RNase P. Cryo-EM structures of RNase MRP and RNase P were resolved at 3.3 Å and 3.0 Å from cell lysates, and 3.6 Å and 3.9 Å from anti-flag elution, respectively. Notably, additional densities were observed in the structures obtained from cell lysates, which were absent in those from the anti-FLAG eluate. These findings establish the GFD-A grid as a robust platform for investigating endogenous proteins, capable of capturing transient interactions and enhancing the resolution of challenging cryo-EM structures with greater efficiency.