The study demonstrates the effective manipulation of antiferromagnetic order in the Weyl semimetal Mn3Sn using orbital torques (OTs) from either metal Mn or oxide CuOx. The critical switching current density is significantly reduced by inserting a heavy metal layer, such as Pt, which enhances the orbital-to-spin conversion efficiency. The memristor-like switching behavior of Mn3Sn mimics the potentiation and depression processes of a synapse with high linearity, making it suitable for constructing accurate artificial neural networks. The work paves the way for manipulating topological antiferromagnetic order and inspires more high-performance antiferromagnetic functional devices.The study demonstrates the effective manipulation of antiferromagnetic order in the Weyl semimetal Mn3Sn using orbital torques (OTs) from either metal Mn or oxide CuOx. The critical switching current density is significantly reduced by inserting a heavy metal layer, such as Pt, which enhances the orbital-to-spin conversion efficiency. The memristor-like switching behavior of Mn3Sn mimics the potentiation and depression processes of a synapse with high linearity, making it suitable for constructing accurate artificial neural networks. The work paves the way for manipulating topological antiferromagnetic order and inspires more high-performance antiferromagnetic functional devices.