Zhipu and MiniMax Each Surge Over 23% on Policy and Model News

On a single trading day in Hong Kong, Bloomberg reported that Chinese AI companies Zhipu and MiniMax each surged at least 23%, extending year-to-date gains that are difficult to contextualize without pausing. Zhipu's is reportedly around 2,000% for the year. MiniMax's is around 260%.

The immediate catalysts were a mix of product news and policy signals. According to the reporting, several Chinese AI companies unveiled upgraded models and agents, with Zhipu announcing GLM-5.2, described as its most capable open-source large model, to be released with no usage restrictions. Top policymakers also renewed calls for broader adoption of AI technology, a signal the market read as continued state-level backing. The Bloomberg piece also noted that Washington's tightening of curbs on foreign access to its most powerful models has shaped the narrative around Chinese AI companies' potential to capture global demand.

Institutional money is arriving with more structure than before. Bank of America analysts initiated coverage on both companies with buy ratings, setting a price target of HK$1,250 for Zhipu and HK$500 for MiniMax. Underpinning that is a concrete structural shift: both companies joined the Hang Seng Tech Index effective June 5, and Bloomberg Intelligence estimates the combined southbound inflows the inclusion could unlock reach HK$139 billion, with Zhipu alone potentially attracting up to HK$92 billion.

The honest caveat is that the reporting does not give you the fundamentals behind these prices. There are no revenue figures, no profitability metrics, no measure of how much enterprise adoption is actually driving valuation versus index mechanics and analyst momentum. A 2,000% year-to-date move demands that question, and the source does not answer it. The reporting also does not address how durable Beijing's policy support looks in concrete terms beyond the renewed public calls cited here.

For anyone tracking where capital is flowing in AI, the forward question is whether GLM-5.2's open-source availability translates into developer adoption at scale, and whether the Hang Seng inclusion creates a sustained bid or a one-time rebalancing pop that fades once passive funds have finished their buying.