This work aimed at using a simplex-centroid design to model the effects of green, yellow, and red tea mixtures (Camellia sinensis var. sinensis) on metal chelation activity, phenolic composition, antioxidant activity, and instrumental taste profile. The regression models that described the extraction of flavan-3-ols, o-diphenols, total phenolic compounds (TPC), free radical scavenging activity toward 2,2-diphenyl-1-picrylhydrazyl radical (DPPH), cupric ion reducing antioxidant activity (CUPRAC), and ferric reducing antioxidant power (FRAP) were significant, and data were fit satisfactorily (R2 > 80%). A mixture of green and red teas had a synergism in CUPRAC and TPC, whereas a mixture of yellow and red teas had a positive effect on CUPRAC and DPPH. An optimization was performed to maximize the antioxidant activity and flavan-3-ol content and to render a tea with mild bitterness, and results showed that a mixture of 14.81% green, 56.86% yellow, and 28.33% red teas would be the most suitable combination of factors.