This preformulation study aims to develop an optial oaral Tirzepaitde table by utilizing AI to elvauat its physicochemical properties and compare them with experimental data. This study aims to evaluate the physicochemical properties of active drug components. Tirzepatide is a dual GLP-1/GIP receptor co-agonist which is used to improve blood sugar control in patients with type 2 diabetes and for the treatment of obesity. We aimed to obtain fundamental data for the optimal formulation design by predicting, comparing, and evaluating the preformulation properties of Tirzepatide through self-developed AI system. We evaluated the solubility of Tirzepatide in 17 solvents and solutions. It exhibited ‘freely soluble’ or ‘soluble’ in Ethanol(EtOH), Methanol(MeOH), and at pH 6.8 to 11.0, and exhibited ‘insoluble’ in Acetonitrile(ACN) and at pH 1.0 to 5.0. These results showed considerable deviation from AI-based predictions and the prediction data for the physicochemical properties of peptide showed limitations. To confirm the interaction between Tirzepatide and 12 excipients, compatibility studies were performed for 4 weeks at a 1:1 ratio under long-term storage conditions(4℃) and accelerated storage conditions(25℃/60%RH). The results indicated that Povidone K90 was incompatible with Tirzepatide. These results are expected to be used as basic data for the design of optimal formulations of Tirzepatide as an oral tablet, and to help improve the analysis of peptides in AI.