What happened
Chinese AI lab Zhipu (Knowledge Atlas Technology) saw its Hong Kong-listed shares surge 48%, closing up 33%, after announcing the open-source release of its GLM-5.2 model this week. This followed US government restrictions forcing Anthropic to withdraw its Claude models from foreign users. GLM-5.2, with a one-million-token context window, performs comparably to Anthropic’s Claude Opus 4.7 on coding and long-horizon agent tasks in preliminary community tests. JPMorgan raised Zhipu's target to HK$1,400 from HK$950; Bank of America initiated buy ratings for Zhipu and MiniMax. Zhipu's market value reached HK$489bn, a tenfold increase since its January IPO.
Why it matters
Access to frontier AI models shifts as US restrictions on Anthropic create a vacuum for non-US enterprises, which Chinese firms like Zhipu are filling with open-source offerings. Procurement teams and platform engineers seeking capable, cost-effective models now have more options, potentially reducing reliance on restricted US providers. This dynamic also creates competitive pressure on standalone vendors and could accelerate "brain flight" of Chinese-born US AI engineers towards labs like DeepSeek and Moonshot, per Z-Ben Advisors. Zhipu's prior 8-17% API price increases indicate emerging pricing power in this segment.
