Electroluminescence (EL) from the singlet-excited (S1) state is the ideal choice for stable, high-performing deep-blue organic light-emitting diodes (OLEDs) owing to the advantages of an adequately short radiative lifetime, improved device durability, and low cost, which are the most important criteria for their commercialization. Herein, we present the design and synthesis of three donor-acceptor-donor (D–A–D)-configured deep-blue fluorescent materials (denoted as TC–1, TC–2, and TC–3) composed of a thioxanthone or diphenyl sulfonyl acceptor and phenyl carbazolyl donor. These systems exhibit strong deep-blue photoluminescence (422–432 nm) in solutions and redshifted emission (472–486 nm) in thin films. The solid-state photoluminescence quantum yield (PLQY) was estimated to be 78 and 94% for TC–2 and TC–3, respectively. TC–2 and TC–3 possess good molecular packing and large molecular cross-sectional areas, which not only improves the PLQY but enhances the triplet–triplet annihilation up-conversion (TTAUC) efficiency of fluorescent emitters. Furthermore, both compounds were applied as an acceptor for confirming their TTAUC property using bis­(2-methyldibenzo­[f,h]­quinoxaline)­(acetylacetonate)­iridium­(III) (Ir­(MDQ)2acac) as the sensitizer. Non-doped OLEDs based on TC–2 and TC–3 exhibit blue EL in the 461–476 nm range. In particular, TC–3 exhibits a maximum external quantum efficiency (EQEmax) of 5.1%, and its EL maximum is 476 nm. In addition, the three emitters were employed as hosts in red OLEDs using bis­(1-phenylisoquinoline)­(acetylacetonate)­iridium­(III) (Ir­(piq)2acac) as the phosphorescent dopant. The red phosphorescent OLEDs based on TC–1, TC–2, and TC–3 achieve excellent EQEmax values of 21.6, 22.9, and 21.9%, respectively, and peak luminance efficiencies of 12.0, 14.0, and 12.3 cd A–1. These results highlight these fluorophores’ versatility and promising prospects in practical OLED applications.