Apex Intelligence Raises Nearly RMB400M for Self-Improving AI Research

Chen Yongchao, founder of AI startup Apex Intelligence, said in an interview that the company has recently closed nearly RMB400 million in angel and angel-plus rounds. The angel round was co-led by IDG Capital, Xinglian Capital and XtalPi, while the angel-plus round was co-led by Zhongguancun Science City Fund, Shenzhen Capital Group and Shanghai Future Industries Fund, with several other investors participating. The funding will mainly be used for foundation model development, compute infrastructure construction, recursive self-improvement research and team expansion.

Chen, a 27-year-old assistant professor at Tsinghua University, said the company's self-improving AI system produced 34 papers in May 2026. Submitted to ACL Rolling Review, 11 papers received scores of 3 or above, and two outperformed 99% of human researchers. He described the result as beyond expectations but said such progress was inevitable.

Apex Intelligence focuses on recursive self-improvement at the foundation model level rather than building agent-layer applications. The system uses research trajectories generated during tasks as training data to improve the model. The company has accumulated tens of thousands of research trajectories, and its AI research system has been launched in a limited trial. Chen said the first model is expected to be released by the end of 2026 or early 2027.