Although pancreatic cancer, particularly PDAC with the KRAS G12D mutation, has been the focus of various studies, research specifically evaluating the potential of active compounds from Uncaria gambir as KRAS G12D inhibitors through an in silico approach remains limited. This study aimed to analyze the binding affinity, drug-likeness profile, and ADMET parameters of active compounds from Uncaria gambir against the KRAS G12D protein. This study used a computational approach through molecular docking and virtual screening designs involving seven test compounds with the KRAS G12D protein structure (PDB ID: 7RPZ), using MRTX1133 as a positive control. The data were analyzed based on binding affinity, RMSD, Lipinski’s Rule of Five, Veber parameters, and ADMET evaluation. The results showed that roxburghine had the highest affinity (−6.7780 kcal/mol), but did not meet Lipinski’s criteria and was indicated to be hepatotoxic. Gambirine and isogambirine were detected as mutagenic and hepatotoxic, whereas quercetin was considered the most prospective because it had a binding affinity of −5.1256 kcal/mol, an RMSD of 1.0570 Å, and interactions with GLU63, HIS95, and GLN99 residues through H-acceptor bonds, accompanied by a superior pharmacokinetic and safety profile. The conclusion of this study confirms that active compounds from Uncaria gambir, particularly quercetin, have the potential to be further explored as candidate KRAS G12D inhibitors in pancreatic cancer, while also providing an initial contribution to the development of natural compounds based on computational approaches in anticancer research.