On February 12, Zhipu released a price adjustment notice for the GLM Coding Plan, stating that recent market demand for the GLM Coding Plan has continued to grow strongly, with user base and call volume increasing rapidly.
To ensure stability and service quality under high load, the company is simultaneously increasing investment in computing power and model optimization, with ongoing product capability upgrades. Based on actual usage and resource input changes, the company has decided to make a structural adjustment to the pricing system of the GLM Coding Plan packages.
The adjustment details are as follows: the initial purchase discount will be canceled, while quarterly and annual subscription discounts will be retained; package prices will be structurally adjusted with an overall increase starting from 30%; prices for existing subscribers will remain unchanged. Effective date: February 12, 2026.
In the secondary market, Zhipu, known as the “First Stock of Large Models” that only listed on the Hong Kong Stock Exchange in January, has recently experienced a strong rally. Its market capitalization once surpassed HKD 150 billion.
Besides Zhipu, many AI companies such as MiniMax have also become popular among Hong Kong investors. As several major tech giants have recently upgraded their large models and more AI applications are being launched, the investment value of AI companies is attracting increasing attention from investors.
The “Twin Giants” of Large Models Are Gaining Popularity
Zhipu’s recent rise began with rumors of a new model launch. The global model service platform OpenRouter recently launched an anonymous model codenamed “Pony Alpha,” which has attracted high attention due to its coding capabilities reaching the top tier worldwide. The market generally believes this new model is Zhipu’s GLM-5. Previously, Zhipu’s chief scientist Tang Jie publicly hinted “GLM is coming soon,” which was interpreted as a teaser before the new model’s release. Securities Times reporters inquired about these rumors with Zhipu but received no response.
“Zhipu has always been relatively low-profile before going public. Its main clients are government and enterprise customers. Whether this anonymous model belongs to Zhipu or not, the company’s R&D capabilities in large models are very strong,” said an analyst who has been tracking domestic large model development. GF Securities also pointed out in its latest research report that Zhipu has built a complete technical matrix centered on its self-developed GLM base model, covering multimodal, agent, and code generation directions, with multiple monetization paths through API services, local deployment, and industry-specific solutions.
“Besides the new model, the scarcity of listed targets in Hong Kong stocks is also a reason for market enthusiasm. Zhipu, as the ‘First Stock of Large Models’ in Hong Kong, along with MiniMax, which also listed in January, are known as the ‘Twin Giants’ of Hong Kong’s large model sector. Their listings fill a gap in Hong Kong’s AI base targets and have become essential assets for global capital allocation. Coupled with market enthusiasm for AI hot topics and short-term speculative capital, this has further driven rapid short-term stock price increases,” the analyst added.
Foreign investment banks are also optimistic about the recent surge in large model companies’ stock prices. JPMorgan analyst Xu Wentao recently published a research report initiating coverage of Zhipu and MiniMax with “Overweight” ratings, with target prices of HKD 400 and HKD 700 respectively, viewing them as top picks to capture the next wave of global AI value creation. The report believes China’s AI industry is entering a new era defined by commercialization, model innovation, and global expansion, with these two companies standing out as the most distinctive and rapidly globalizing independent large model developers.
However, a private equity fund manager in Shenzhen, who invests in Hong Kong stocks, warned of risks regarding the sharp recent stock price increases: “There have been many cases where newly listed stocks were heavily speculated upon after listing, mainly to meet the criteria for inclusion in major stock indices and further into the Hong Kong Stock Connect. But afterwards, stock prices are likely to pull back.”
Intensive Launches of AI Applications
In fact, the recent upgrades of new models by major tech giants and the launch of new applications have greatly increased investor attention on the AI sector.
Securities Times reporters learned that leading domestic tech giants such as ByteDance and Alibaba have announced intensive updates around the Spring Festival. ByteDance recently released the AI video generation model Seedance 2.0, which supports text or image input and can generate multi-angle videos with native audio within 60 seconds; Alibaba’s new flagship AI model Qwen 3.5 is scheduled for release during the Spring Festival; DeepSeek, after a year, also plans to launch its new generation large model during the same period. On the product front, Yuanbao, Doubao, and Qwen are competing in the Spring Festival AI red envelope battle. Overseas, open-source project Clawdbot has gained attention for its ability to run locally on Mac or servers to serve as a personal AI assistant, quickly becoming popular.
“In the short term, the underlying models will not undergo fundamental changes; the main improvements will focus on reducing costs and enhancing coherence. In the long term, models that can operate in real-world scenarios will become a trend,” said Zhang Renqi, Executive Director of the Investment Department at Cornerstone Capital. Cornerstone Capital had already invested in MiniMax in 2023 and has deep research and investments in AI.
However, despite industry rapid growth, there are still multiple challenges, especially high investment and high loss issues, with most AI companies still in the “burning money” stage.
“Commercialization is the current focus. With Zhipu and MiniMax listed in Hong Kong, the large model industry has entered a validation phase in the capital market. The industry has moved beyond reckless growth into a mature stage emphasizing both models and products, as well as general and vertical applications. Vertical applications will be key growth points—scenarios like social entertainment and emotional companionship have proven demand, with stable audiences, even if not yet essential; meanwhile, professional fields like healthcare and finance are also hot. For example, in healthcare, this year’s market is very active, with companies like Ant Fortune and Baichuan making strategic layouts,” said Zhang.
The Era of Global Competition Begins
“From the perspective of industry competition, the rapid iteration period of ‘daily model updates’ from 2023 to 2024 has passed, and the growth in training model technology has slowed. Currently, the domestic market has basically formed a competition pattern of several large companies plus several leading independent firms,” Zhang said about future prospects.
JPMorgan’s research report also states that China’s AI industry is shifting from the “Hundred Models Battle” phase to a stage where success depends on commercialization, model innovation, and global deployment. The Chinese AI market is rapidly consolidating, with the number of capable and well-funded model developers shrinking from over 200 to fewer than 10. JPMorgan further notes that the largest profit pools in China’s AI industry may flow to platform giants with distribution capabilities; independent firms’ breakthroughs depend on finding niche survival spaces through “structural neutrality.”
In the global AI competition, US tech giants continue to invest heavily. Amazon plans to spend $200 billion annually, Google expects $175-185 billion, Meta $115-135 billion, and Microsoft over $100 billion, with all surpassing market expectations significantly.
“Chinese companies are strong in commercialization and are expected to hold a significant share in the global open-source model market, laying a foundation for international expansion. The ‘coming of age’ for large models has just begun, and more industry transformations are brewing,” Zhang concluded.
(Article source: Securities Times)
View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
The first major model company officially announces: Price increase! Multiple major companies are intensively upgrading to new models!
Zhipu Announces Price Increase
On February 12, Zhipu released a price adjustment notice for the GLM Coding Plan, stating that recent market demand for the GLM Coding Plan has continued to grow strongly, with user base and call volume increasing rapidly.
To ensure stability and service quality under high load, the company is simultaneously increasing investment in computing power and model optimization, with ongoing product capability upgrades. Based on actual usage and resource input changes, the company has decided to make a structural adjustment to the pricing system of the GLM Coding Plan packages.
The adjustment details are as follows: the initial purchase discount will be canceled, while quarterly and annual subscription discounts will be retained; package prices will be structurally adjusted with an overall increase starting from 30%; prices for existing subscribers will remain unchanged. Effective date: February 12, 2026.
In the secondary market, Zhipu, known as the “First Stock of Large Models” that only listed on the Hong Kong Stock Exchange in January, has recently experienced a strong rally. Its market capitalization once surpassed HKD 150 billion.
Besides Zhipu, many AI companies such as MiniMax have also become popular among Hong Kong investors. As several major tech giants have recently upgraded their large models and more AI applications are being launched, the investment value of AI companies is attracting increasing attention from investors.
The “Twin Giants” of Large Models Are Gaining Popularity
Zhipu’s recent rise began with rumors of a new model launch. The global model service platform OpenRouter recently launched an anonymous model codenamed “Pony Alpha,” which has attracted high attention due to its coding capabilities reaching the top tier worldwide. The market generally believes this new model is Zhipu’s GLM-5. Previously, Zhipu’s chief scientist Tang Jie publicly hinted “GLM is coming soon,” which was interpreted as a teaser before the new model’s release. Securities Times reporters inquired about these rumors with Zhipu but received no response.
“Zhipu has always been relatively low-profile before going public. Its main clients are government and enterprise customers. Whether this anonymous model belongs to Zhipu or not, the company’s R&D capabilities in large models are very strong,” said an analyst who has been tracking domestic large model development. GF Securities also pointed out in its latest research report that Zhipu has built a complete technical matrix centered on its self-developed GLM base model, covering multimodal, agent, and code generation directions, with multiple monetization paths through API services, local deployment, and industry-specific solutions.
“Besides the new model, the scarcity of listed targets in Hong Kong stocks is also a reason for market enthusiasm. Zhipu, as the ‘First Stock of Large Models’ in Hong Kong, along with MiniMax, which also listed in January, are known as the ‘Twin Giants’ of Hong Kong’s large model sector. Their listings fill a gap in Hong Kong’s AI base targets and have become essential assets for global capital allocation. Coupled with market enthusiasm for AI hot topics and short-term speculative capital, this has further driven rapid short-term stock price increases,” the analyst added.
Foreign investment banks are also optimistic about the recent surge in large model companies’ stock prices. JPMorgan analyst Xu Wentao recently published a research report initiating coverage of Zhipu and MiniMax with “Overweight” ratings, with target prices of HKD 400 and HKD 700 respectively, viewing them as top picks to capture the next wave of global AI value creation. The report believes China’s AI industry is entering a new era defined by commercialization, model innovation, and global expansion, with these two companies standing out as the most distinctive and rapidly globalizing independent large model developers.
However, a private equity fund manager in Shenzhen, who invests in Hong Kong stocks, warned of risks regarding the sharp recent stock price increases: “There have been many cases where newly listed stocks were heavily speculated upon after listing, mainly to meet the criteria for inclusion in major stock indices and further into the Hong Kong Stock Connect. But afterwards, stock prices are likely to pull back.”
Intensive Launches of AI Applications
In fact, the recent upgrades of new models by major tech giants and the launch of new applications have greatly increased investor attention on the AI sector.
Securities Times reporters learned that leading domestic tech giants such as ByteDance and Alibaba have announced intensive updates around the Spring Festival. ByteDance recently released the AI video generation model Seedance 2.0, which supports text or image input and can generate multi-angle videos with native audio within 60 seconds; Alibaba’s new flagship AI model Qwen 3.5 is scheduled for release during the Spring Festival; DeepSeek, after a year, also plans to launch its new generation large model during the same period. On the product front, Yuanbao, Doubao, and Qwen are competing in the Spring Festival AI red envelope battle. Overseas, open-source project Clawdbot has gained attention for its ability to run locally on Mac or servers to serve as a personal AI assistant, quickly becoming popular.
“In the short term, the underlying models will not undergo fundamental changes; the main improvements will focus on reducing costs and enhancing coherence. In the long term, models that can operate in real-world scenarios will become a trend,” said Zhang Renqi, Executive Director of the Investment Department at Cornerstone Capital. Cornerstone Capital had already invested in MiniMax in 2023 and has deep research and investments in AI.
However, despite industry rapid growth, there are still multiple challenges, especially high investment and high loss issues, with most AI companies still in the “burning money” stage.
“Commercialization is the current focus. With Zhipu and MiniMax listed in Hong Kong, the large model industry has entered a validation phase in the capital market. The industry has moved beyond reckless growth into a mature stage emphasizing both models and products, as well as general and vertical applications. Vertical applications will be key growth points—scenarios like social entertainment and emotional companionship have proven demand, with stable audiences, even if not yet essential; meanwhile, professional fields like healthcare and finance are also hot. For example, in healthcare, this year’s market is very active, with companies like Ant Fortune and Baichuan making strategic layouts,” said Zhang.
The Era of Global Competition Begins
“From the perspective of industry competition, the rapid iteration period of ‘daily model updates’ from 2023 to 2024 has passed, and the growth in training model technology has slowed. Currently, the domestic market has basically formed a competition pattern of several large companies plus several leading independent firms,” Zhang said about future prospects.
JPMorgan’s research report also states that China’s AI industry is shifting from the “Hundred Models Battle” phase to a stage where success depends on commercialization, model innovation, and global deployment. The Chinese AI market is rapidly consolidating, with the number of capable and well-funded model developers shrinking from over 200 to fewer than 10. JPMorgan further notes that the largest profit pools in China’s AI industry may flow to platform giants with distribution capabilities; independent firms’ breakthroughs depend on finding niche survival spaces through “structural neutrality.”
In the global AI competition, US tech giants continue to invest heavily. Amazon plans to spend $200 billion annually, Google expects $175-185 billion, Meta $115-135 billion, and Microsoft over $100 billion, with all surpassing market expectations significantly.
“Chinese companies are strong in commercialization and are expected to hold a significant share in the global open-source model market, laying a foundation for international expansion. The ‘coming of age’ for large models has just begun, and more industry transformations are brewing,” Zhang concluded.
(Article source: Securities Times)