Today, the exploitation of recycled fibers emanating from textile waste has substantially increased owing to the growing environmental consciousness, legal sustainability mandates, and escalating raw material expenses. Herein, for the very first time, an innovative and sustainable approach was employed to utilize the dyed waste jute fibers accumulated beneath the jute weaving machines to create mélange yarns. Recycled cotton was used as well to reduce the amount of virgin cotton in mélange yarn. First-time, a precise prediction of mixing proportions in mélange yarn was yielded by using an I-optimal mixture design, a part of response surface methodology. The design recommended performing 16 experimental runs with 3 mixing components: dyed waste jute fiber (5–30%), recycled cotton (5–15%) and virgin cotton (55–80%), resulting in the production of 16 distinct yarn samples. Surface response models such as special quartic, special cubic, linear, and quadratic were developed for predicting the yarn unevenness, imperfection, hairiness, strength, and elongation. All the models were found to be statistically significant with p-values below 0.0500. Using desirability functions based on the response models, the optimized mixture components in the yarn were derived to be 15.8% dyed waste jute, 7.4% recycled cotton, and 76.8% virgin cotton fibers. This formulation was used to produce an 18 Ne mélange yarn whose quality parameters closely matched the predicted values, substantiating the effectiveness of the response model in navigating and predicting the attributes of mélange yarn.