MiniMax CEO Yan Junjie Faces Heightened Selling Pressure as Lock‑up Shares Near Expiration After $39 B Market Decline

MiniMax Group founder Yan Junjie has lost roughly $9.3 billion in personal wealth as the Hong Kong‑listed AI model developer shed about $39 billion in market value since its March peak. The 37‑year‑old chairman and CEO’s net worth has dropped to $3.3 billion from a high of $12.6 billion, according to Forbes estimates. The share price has plunged more than 70 % from a record HK$1,330 (≈$170) hit in March after a $619 million IPO in January.

The stock is up 1.6 % year‑to‑date, yet investor sentiment has turned bearish. MiniMax’s latest M3 model shows advances in coding and complex agentic tasks, but competitors such as Zhipu (Z.AI) are perceived as moving faster, keeping MiniMax at a domestic disadvantage.

“The market is pessimistic on MiniMax because the company is now at a disadvantage domestically,” says Charlie Chai, a Shanghai‑based analyst at 86Research.

The company launched the M3 model in June, touting “frontier‑level performance on specialized tasks such as coding and agentic work.” It also claims significant improvements over the previous M2 model, especially in debugging generated code.

Meanwhile, investors have rotated capital toward Zhipu, Hong Kong’s other major listed AI developer. Zhipu’s stock has surged nearly 1,300 % this year, propelling co‑founder and Chairman Liu Debing to the 8th richest person in China with a net worth of $28.4 billion, while fellow co‑founder Tang Jie is also a billionaire at $6.1 billion, according to Forbes’ Real‑Time Billionaires List.

Tang recently used X (formerly Twitter) to highlight Zhipu’s progress, stating that “it won’t take that long” for Chinese models to catch up with the best U.S. offerings, countering Elon Musk’s forecast that Chinese AI could match Anthropic’s Claude Fable by early 2027.

Zhipu’s GLM 5.2 model, also released in June, has drawn attention in Silicon Valley. Despite being on a U.S. trade blacklist, the model is claimed to excel in long‑horizon tasks and to outperform OpenAI’s GPT‑5.5 in coding while lagging slightly behind Anthropic’s Claude Opus 4.8. The company says GLM 5.2 is the first Chinese model to match or surpass major American AI models without compromise.

“Many smart people/AI insiders are saying GLM-5.2 is the first Chinese AI model to match and often beat the American big lab public AI models with no compromises,” U.S. billionaire investor Marc Andreessen wrote in a June 28 post via the X social media platform.

MiniMax faces additional pressure as lock‑up restrictions on a portion of its publicly traded shares are set to expire on Wednesday. Approximately 153.5 million extra shares will become tradable, raising the free float to over 50 % from 4.1 % according to a June note from DZT Research. “We do not yet see MiniMax proving its edge against DeepSeek and GLM,” Ke Yan, Singapore‑based head of research at DZT Research, wrote in the note. “With the model still unproven and the lock‑up supply ahead, we are bearish on the stock.”

Zhipu will also see about 25.7 million shares released from lock‑up on Tuesday, though more than 90 % of its listed shares remain restricted until next January, per 86Research’s Chai.

Alibaba Group Holding, a key cornerstone investor in MiniMax, has voiced confidence in the firm’s technology. An Alibaba spokesperson confirmed the company’s bullish view when contacted by Forbes, noting willingness to collaborate on cloud computing and enterprise services.

BofA Securities analyst Alex Liu sees a potential “catch‑up trade” for MiniMax. He highlighted the wide valuation gap—MiniMax trades at 19× estimated 2027 sales versus Zhipu’s 52× forward multiple—and noted the possibility of MiniMax being added to the Stock Connect Southbound link in early August, which could bolster demand.

Source link