Why are Chinese AI stocks surging?

Investor money poured into Chinese artificial intelligence companies Thursday after several businesses rolled out new technology and Beijing’s leadership called for faster adoption across the country.

Zhipu AI, which is listed on the Hong Kong exchange under the name Knowledge Atlas Technology, jumped almost 30% to close at 405 Hong Kong dollars ($51.8). The spike happened right after the company put out GLM-5, a model anyone can use that handles coding work and manages tasks over long stretches without stopping.

Zhipu said its new tool performs close to Anthropic’s Claude Opus 4.5 when tested on coding tasks and beats Google’s Gemini 3 Pro in certain areas, though no one outside the company has checked those numbers yet.

MiniMax shares climbed 13.7% to reach 70.5 Hong Kong dollars by market close after launching an updated version called M2.5 on Wednesday. The company markets this as “a model built for Max coding & agentic workflows” with better tools for AI agents, released through its international website.

The market reaction to these announcements shows investors remain confident in Chinese AI businesses despite global uncertainty. UCloud Tech, a Shanghai-listed firm that provides computing power for Zhipu, hit its daily trading limit with a 20% gain on Thursday.

SenseTime, which changed its focus from facial recognition surveillance equipment to AI software platforms, saw its Hong Kong shares rise 6.8%.

DeepSeek, the startup that grabbed worldwide attention and caused major market disruptions last year, updated its main AI model on Wednesday by expanding what it can process and adding fresher information. The company has faced growing regulatory scrutiny from governments concerned about data security and its connections to Chinese authorities.

Ant Group released its Ming-Flash-Omni 2.0 model Wednesday, described as a tool that can create speech, music, sound effects and images. ByteDance kicked off the week Monday by launching Seedance 2.0, its newest AI video tool, which triggered gains across Chinese AI app stocks.

Reports suggest the company is also working with Samsung to develop its own chips.

Beijing signals strong government support

Premier Li Qiang stressed Wednesday that China needs to push for “scaled and commercialized application of AI” while improving coordination of computing resources and power systems. He emphasized Beijing’s plans to create a better environment for AI talent and businesses.

The surge in pure AI startups contrasts with drops in major Chinese tech companies that also run AI divisions. Tencent shares fell 2.6% while Alibaba declined 2.1%. The Hong Kong Hang Seng Tech index dropped 1.7%.

These developments happen as Chinese companies race to match American competitors while spending far less money. A RAND research group report published last month found Chinese models run at roughly one-sixth to one-fourth the cost of similar American systems. This approach lets Chinese firms focus on practical uses rather than chasing huge capital spending.

Tai Hui, APAC chief market strategist at JPMorgan, told CNBC Thursday he thinks talk of an AI bubble seems “a little premature” since many quality companies globally still have solid fundamentals.

JPMorgan analysts recently started coverage of Zhipu and MiniMax with “Buy” ratings, projecting MiniMax will see revenue grow at 138% annually through 2030 while Zhipu grows 127% over the same period, with both companies expected to turn profitable in 2029.

Rolf Bulk, an analyst at Futurum Group, noted Chinese technology companies have taken a relatively frugal approach to AI development, with far less capital spending than American rivals while focusing on the domestic market.

As Cryptopolitan previously covered, DeepSeek’s breakthrough showed the industry that powerful models can be created even under resource constraints, forcing Chinese competitors to accelerate their own development timelines.

Chinese AI firms are preparing for sustained competition with US companies. With government backing and lower operating costs, these businesses might capture growing market share in developing countries where affordability matters more than cutting-edge features.

However, ongoing regulatory concerns about data security and government access could limit their expansion into Western markets.