The rapid development of technologies such as machine learning and artificial intelligence has further accelerated the growth of multi-value logic (MVL) computing nanodevices. However, current MVL nanodevices based on conventional materials face the challenges of low peak−valley ratio (PVR) and unstable spin filtering effects. Therefore, the intrinsic half-metal nanoscale MXene has become the answer to the development bottleneck of MVL nanodevices for its outstanding properties such as high Curie temperature and an appropriate band structure. The present work attempts to design spintronic magnetic tunnel junctions (MTJs) based on new predicted Sc 2 CHO as half-metallic electrodes and Sc 2 NHO as semiconductor-based scattering regions. Furthermore, we simulate its transport properties from density functional theory combined with nonequilibrium Green's function. And we have considered the effect of the length of the scattering region on MTJ performance. According to our calculations, the spintronic transport of the MTJs demonstrates up to perfect 100% polarizability spin currents stably due to the single-channel conduction capability of half-metal Sc 2 CHO. The MTJs exhibit a significant negative differential resistance (NDR) phenomenon, with the PVR of the type 2N reaching 1.28 × 10 6 , the highest value among the two-dimensional spintronic nanodevices currently being theoretically explored. MTJs with a high PVR will be beneficial for the application and development of MVL nanodevices. The spin filtering effect and NDR phenomenon can be well-maintained as the scattering region length increases, and the PVR usually almost always remains 10 4 . This is well reflected in the transmission spectra and molecular projection self-consistent Hamiltonian. The tunneling magnetoresistance phenomenon is also observed.