Hexagonal-Boron nitride nanotubes (h-BN) decorated with transition metals have been widely studied due to their enhanced physicochemical properties. In this paper, Pt cluster-modified h-BN is proposed as a sensitive material for a novel gas sensor for the online malfunction monitoring of oil-immersed transformers. The inner oil is ultimately decomposed to various gases during the long-term use of oil-immersed transformers. Exposure to excessively high temperatures produces the alkanes CH4 and C2H6, whereas different degrees of discharge generate H2 and C2H2. Therefore, the identification of H2, CH4, and C2H2 gas efficiently measures the quality of transformers. Based on the density functional theory, the most stable h-BN doped with 1–4 Pt atoms is employed to simulate its adsorption performance and response behavior to these typical gases. The adsorption energy, charge transfer, total density of states, projected density of states, and orbital theory of these adsorption systems are analyzed and the results show high consistency. The adsorption ability for these decomposition components are ordered as follows: C2H2 > H2 > CH4. Pt cluster-modified h-BN shows good sensitivity to C2H2, H2, with decreasing conductivity in each system, but is insensitive to CH4 due to its weak physical sorption. The conductivity change of Ptn-h-BN is considerably larger upon H2 than that upon C2H2, but is negligible upon CH4. Our calculations suggest that Pt cluster modified h-BN can be employed in transformers to estimate their operation status.