Cancer remains a major global health challenge, and computational approaches can support the early-stage identification of peptide candidates for experimental evaluation. This study aimed to identify putative anticancer peptides (ACPs) encrypted within Lupinus albus β- and γ-conglutins using an in silico screening strategy. Protein sequences from two β-conglutins and two γ-conglutins were subjected to simulated enzymatic hydrolysis with pepsin, trypsin, and chymotrypsin. The resulting peptides were evaluated using PeptideRanker for bioactivity prediction, AntiCP 2.0 for anticancer peptide prediction, ToxinPred 3.0 for toxicity assessment, and ProtParam for physicochemical characterization. Simulated hydrolysis generated 1610 peptide fragments, of which 749 met the length criteria for downstream analysis. PeptideRanker identified 124 peptides with predicted bioactive potential, and AntiCP 2.0 classified 47 as putative ACPs. Toxicity screening reduced this number to 31 peptides predicted as non-toxic, resulting in 18 unique putative non-toxic ACP candidates after redundancy removal. Among these, 16 met the minimum sequence length required for complete ProtParam characterization. These findings indicate that L. albus conglutins may contain encrypted peptide sequences computationally prioritized as putative ACP candidates, whose activity, selectivity, toxicity, and safety require in vitro and/or in vivo validation.