Nanjie Intelligence, a startup company associated with Nanjing University, has unveiled a technical report, introducing its innovative model swarm intelligence technology framework, which is constructed upon extensive model collaboration. The report posits that models constitute the fourth fundamental element for intelligent growth, supplementing computing power, algorithms, and data. This notion is derived from the learning artifact paradigm put forth by Academician Zhihua Zhou. The core system of this framework, Mio 1.0 Lite, emerged as the top performer on average across nine public benchmark tests, all the while achieving the lowest cost per request. This accomplishment substantially cuts down on usage expenses without compromising task performance. Rather than developing general-purpose large models, Nanjie Intelligence envisions itself as the 'power grid' of the model era. It has established a comprehensive system encompassing model repositories, capability characterization, intelligent routing, and scheduling. Presently, its product platform has archived 484 models and has commenced public testing, having successfully concluded its inaugural funding round, which garnered nearly 100 million yuan.