#China #Beijing #funding #AI Research #Apex Intelligence

Apex Intelligence Secures Significant Funding

Apex Intelligence, a Beijing-based startup, has successfully completed its angel and angel-plus funding rounds, amassing approximately $50 million in total. This funding is poised to propel their vision of creating self-evolving foundation models that transform the landscape of artificial intelligence (AI).

A New Paradigm in AI Research

Founded in June 2026, Apex Intelligence is pioneering the concept of Recursive Self-Improvement (RSI), which aims to move AI beyond mere assistance to becoming an active co-researcher. This shift is vital for fostering scientific research and discovery, and Apex believes that the future lies in AI that can autonomously explore the unknown rather than just enhance existing knowledge.

Under the leadership of Yongchao Chen, an assistant professor at Tsinghua University's Institute for AI, the company plans to engage with academia and attract top talent by visiting several prestigious universities in North America, including Harvard and MIT, between September 17 and 23.

The Focus on Self-Evolution

Apex Intelligence envisions a shift in the role AI plays in research and development (R&D). Historically, advancements in technology have corresponded with breakthroughs in scientific research, which often rely on human intellect. However, Apex contends that AI's role should evolve from merely accelerating known processes to actively discovering new pathways in various fields, including chip design and molecular research.

The ambition is to cultivate AI systems capable of generating creative solutions through various phases of experimentation, thereby redefining the R&D process. By harnessing mid-training and post-training techniques, the company aims to guide models through a self-improvement cycle comprising hypothesis generation, experimentation, validation, and iterative learning.

The Aspirational Goal

Long term, Apex Intelligence aspires to develop AI researchers that can collectively surpass a network of 100,000 top-tier scientists across multiple disciplines. This objective is not just an iterative enhancement; it represents a radical transformation in research paradigms, with the potential to revolutionize fields requiring expert input and costly trial-and-error methods.

Notably, their AI system has demonstrated promising results in both AI and mathematical endeavors. In the realm of AI for AI, it has produced research levels comparable to those achieved by PhD students in AI at leading conferences. Furthermore, it has reached new benchmarks in several projects, outpacing developments made by significant players in the industry, including Tencent and Stanford.

In addition, their mathematical applications include solving longstanding conjectures and delivering breakthroughs in various theoretical domains, showing that these self-evolving models can produce results previously thought unattainable.

The Vision Forward

As Apex Intelligence continues to refine its foundation models, it plans to enhance its research trajectory data infrastructure and maintain a focus on recruiting world-class talent. The company's mission is encapsulated in its guiding principle: "Unlock Undiscovered Discovery."

Apex's founder, Yongchao Chen, highlights a fundamental belief: The intelligence of AI systems trained by humans is inevitably bound by human limitations. He insists that true intelligence must align with the objective world, setting the stage for AI systems to take on roles traditionally reserved for humanity's most skilled scientists.

In summary, Apex Intelligence stands at a critical juncture, seeking to redefine the boundaries of AI within the research landscape. With substantial financial backing and a clear vision, the company is poised to usher in a new era of discovery where AI systems evolve to become autonomous researchers, fundamentally changing the way we perceive scientific inquiry.