The capital injection arrives in two tranches. An initial angel round saw backing from IDG Capital, LinkX Capital, and XtalPi, while a subsequent angel-plus round was co-led by Zhongguancun Science City Fund, SCGC, and Shanghai Engine Fund. These resources will fuel the company’s push toward Recursive Self-Improvement (RSI), a framework where models iterate through hypothesis, experimentation, and validation without human intervention.

Unlike standard large language models that optimize for human preference, Chen argues that current architectures are capped by human knowledge. Apex Intelligence seeks to bypass this ceiling by training models to navigate the objective world. The company’s system has already demonstrated early success, producing research papers at the level of a PhD student and solving the 30-year-old majorization conjecture in mathematics. The team is now initiating a recruitment tour across North American institutions, including Harvard, MIT, and Yale, to scale its research infrastructure. The ultimate objective is to create a network of AI researchers whose collective output mirrors that of 100,000 top-tier human scientists, targeting high-stakes industries like molecular R&D and chip architecture.