This article presents a quantum annealing-aided design method for ultrathin metamaterial optical diodes. The researchers developed an efficient design scheme using quantum annealing (QA) and active learning to automatically identify optimal designs for optical diodes at specific wavelengths. The optical diode is a stratified volume diffractive film discretized into rectangular pixels, each assigned to either a metal or dielectric material. The proposed method maximizes the quality of optical isolation at given wavelengths by identifying the optimal material states of each pixel. The study successfully identified optimal structures for three specific wavelengths (600, 800, and 1000 nm). In the best-case scenario, when the forward transmissivity is 85%, the backward transmissivity is 0.1%. Electromagnetic field profiles reveal that the designed diode strongly supports surface plasmons coupled across counterintuitive metal–dielectric pixel arrays, resulting in high amplitude transmission of first-order diffracted light. In contrast, backward transmission has decoupled surface plasmons that redirect Poynting vectors back to the incident medium, resulting in near attenuation of its transmission. The researchers experimentally validated the optical isolation function of the optical diode. The design process involved discretizing the stratified volume diffractive film into rectangular pixels and using a surrogate model based on supervised machine learning to map the design into a binary optimization problem compatible with QA. The QA-enhanced active learning scheme was used to optimize the design, allowing the identification of optimal material states for each pixel. The method was tested with different pixel counts (N = 12, 16, 20, 24, 32, 40) and showed high performance for N = 32 at 800 nm. The study demonstrates that quantum annealing can efficiently solve complex optimization problems in the design of optical diodes, leading to high optical isolation performance. The results show that the proposed method can significantly improve the design efficiency and performance of optical diodes, making them more suitable for miniaturized photonic systems. The experimental validation confirms the effectiveness of the designed optical diodes, highlighting the potential of quantum annealing in optical device design.This article presents a quantum annealing-aided design method for ultrathin metamaterial optical diodes. The researchers developed an efficient design scheme using quantum annealing (QA) and active learning to automatically identify optimal designs for optical diodes at specific wavelengths. The optical diode is a stratified volume diffractive film discretized into rectangular pixels, each assigned to either a metal or dielectric material. The proposed method maximizes the quality of optical isolation at given wavelengths by identifying the optimal material states of each pixel. The study successfully identified optimal structures for three specific wavelengths (600, 800, and 1000 nm). In the best-case scenario, when the forward transmissivity is 85%, the backward transmissivity is 0.1%. Electromagnetic field profiles reveal that the designed diode strongly supports surface plasmons coupled across counterintuitive metal–dielectric pixel arrays, resulting in high amplitude transmission of first-order diffracted light. In contrast, backward transmission has decoupled surface plasmons that redirect Poynting vectors back to the incident medium, resulting in near attenuation of its transmission. The researchers experimentally validated the optical isolation function of the optical diode. The design process involved discretizing the stratified volume diffractive film into rectangular pixels and using a surrogate model based on supervised machine learning to map the design into a binary optimization problem compatible with QA. The QA-enhanced active learning scheme was used to optimize the design, allowing the identification of optimal material states for each pixel. The method was tested with different pixel counts (N = 12, 16, 20, 24, 32, 40) and showed high performance for N = 32 at 800 nm. The study demonstrates that quantum annealing can efficiently solve complex optimization problems in the design of optical diodes, leading to high optical isolation performance. The results show that the proposed method can significantly improve the design efficiency and performance of optical diodes, making them more suitable for miniaturized photonic systems. The experimental validation confirms the effectiveness of the designed optical diodes, highlighting the potential of quantum annealing in optical device design.