Summary
Chinese AI-focused equities climbed on Friday as investors responded to fresh model releases from two high-profile startups. The gains were concentrated in newly listed AI names and related chipmakers, while the wider technology sector in Hong Kong weakened alongside losses in U.S. peers.
Market moves and company updates
Zhipu, which trades as Knowledge Atlas Tech Joint Stock (HK:2513), rose more than 20% to an intraday record of HK$492.0 after unveiling GLM-5, an open-source large-language model on Thursday. The company said GLM-5 performed at a level comparable to Anthropic's Claude Opus 4.5 in coding tests. Zhipu also announced it was increasing subscription prices for its GLM coding plan by up to 30%, citing a rapid increase in demand.
MiniMax Group Inc (HK:0100) was another standout, climbing over 15% to a record HK$688.0 on Friday after releasing its updated M2.5 open-source AI model on Thursday. MiniMax said the M2.5 release included enhanced agentic tools. Both Zhipu and MiniMax are among China’s so-called AI Tigers and were the first of that cohort to list publicly. The two firms have recorded strong performance since their listings in late-2025.
Spillover across AI-linked supply chain
Investor enthusiasm tied to the new launches supported a broader rally among AI-related chipmakers. Shanghai Iluvatar CoreX SemiCon Co (HK:9903) jumped more than 12%, while GigaDevice Semiconductor Inc (HK:3986) and OmniVision Integrated Circuits (HK:0501) gained 3.1% and 0.6%, respectively. Many of the AI-linked names that outperformed were among those that listed in late-2025.
Context within the wider market
The surge in AI stocks occurred against a backdrop of weakness in the wider technology sector. The tech-heavy Hang Seng index fell 1.6% on Friday, as Chinese technology shares moved lower in tandem with overnight losses among U.S. peers. AI-related issues largely avoided that broader rout during the session.
Implications
The session highlighted concentrated investor interest in firms that have delivered visible product upgrades and commercial signals, such as model releases and pricing changes. Gains were focused on both the application layer - the listed AI startups - and on semiconductor suppliers that provide critical hardware for model development and deployment.