China's AI stock selloff erases $150 billion amid fierce market rivalry

Shares of major Chinese artificial intelligence developers MiniMax Group Inc. and Z.AI Co., also known as Zhipu, have dropped by a combined $150 billion from their peak values. This sharp decline follows their popular market debuts in January. MiniMax fell below its original IPO price on Thursday, erasing a massive 650% gain reached in March. Meanwhile, larger rival Z.AI has dropped 81% from its high point in June. New product rollouts have failed to stop the falling prices. MiniMax launched its M3.1 Flash model on Monday with lower token costs and better performance, but the stock received no boost. Zhipu faced a similarly cold response after releasing its GLM-5.3-FlashX model in September. Data from the Artificial Analysis Intelligence Index shows that both companies lag behind US model leaders in overall intelligence. They also trail behind local competitors like DeepSeek, Alibaba Group Holding Ltd., and Xiaomi Corp. as a flood of new products compete for users. Bank of America analyst Alex Liu noted that China has more than 10 AI labs, but the market will likely shrink to fewer through consolidation as it remains difficult to predict future survivors. Both firms are currently unprofitable, and global price wars make financial recovery harder. JPMorgan Chase & Co. strategists, including Chloe Liu, pointed out that MiniMax earns most of its money outside China, while Zhipu's application programming interface is growing popular with global developers due to strong coding skills. The bank favors Z.AI over MiniMax for this reason. Additional pressure comes from upcoming share lockup expiries in January. Pre-IPO investors are set to unlock about 178 million Z.AI shares and 25 million MiniMax shares for trading, according to Bloomberg data.

Context

Share lockup expiries often put downward pressure on stock prices because early investors may choose to sell their shares as soon as they are allowed. The intense competition and lack of profits reflect a broader trend in fast-growing tech sectors, where too many companies chase market share before building a stable business model.