Abstract: Currently, the dye production industry commonly employs manual color matching, which is characterized by low production efficiency, poor precision and high labor consumption. In addition, the traditional computer color matching techniques necessitate fabric dyeing, along with determination of the apparent depth of dye on fabrics to obtain the formulation. This process is time-consuming, costly, and leads to resource wastage and pollutant generation. Hence, a digital full-wavelength inversion color matching model was proposed in this paper. This model was based on K-M theory and directly conducted full-wavelength measurements and inversions of the target dye and its raw materials in the liquid phase, eliminating the fabric dyeing process that typically lasts for several hours and achieving a full-wavelength response within seconds. The formulation design utilized C.I. Disperse Blue 291∶1, C.I. Disperse Orange 288, and C.I. Disperse Violet 93∶1 as raw materials, with C.I. Disperse Black SRN as target. The results indicated that the dye formulation can be obtained through a single inversion, with the model fitting determination coefficient (R²) and average relative error (MRE) reaching 99.99% and 0.062% respectively. After eliminating interference factors, the color difference (△E) of the fabric verification was less than 0.4, and the intensity reached 100%. This full-wavelength inversion color matching method can rapidly compound medium and high-end dye products, effectively support the enhancement of the automation and digitalization levels of dye production, and avoid the generation of a large quantity of polluted wastewater. It possessed the attributes of being green and low-cost, conforming to the national strategy of cleaner production.