The global artificial intelligence race is moving faster than ever. New models are being released almost every month, benchmark records are constantly changing, and competition among AI companies is becoming increasingly intense.
While names such as Kimi, Qwen, DeepSeek, and Tencent Hunyuan continue to attract attention with new model releases and open-weight strategies, another major Chinese AI company is experiencing a very different trend in the capital market.
That company is Zhipu AI, also known as Z.ai, the developer behind the GLM (General Language Model) series.
Once considered one of China’s most promising AI startups, Zhipu AI reportedly saw its stock decline by approximately 67% from a recent peak. The sharp drop has raised an important question: Is Zhipu AI losing its technological edge, or is the market simply re-evaluating the expectations surrounding AI companies?
From Tsinghua University to China’s AI Race
Zhipu AI was founded in 2019 under the name Beijing Zhipu Huazhang Technology. The company emerged from the Knowledge Engineering Group (KEG) Lab at Tsinghua University, one of China’s most respected institutions for computer science and artificial intelligence research.
The company was founded by Tang Jie, a professor in Tsinghua University’s Department of Computer Science, and Zhang Peng, the company’s CEO and a former researcher at the KEG Lab.
Unlike many AI startups that entered the market after the global success of ChatGPT, Zhipu AI had a strong academic and research foundation long before generative AI became a mainstream trend.
The team worked on the GLM model family and later introduced several important models, including GLM-130B, ChatGLM, GLM-4, and more recently, GLM-5.2.
Its connection to Tsinghua University gave Zhipu AI strong credibility and helped it become one of the most closely watched AI labs in China. The company was often mentioned alongside other major players such as DeepSeek, Moonshot AI, Alibaba’s Qwen, and MiniMax.
A Rising Star During the AI Boom
As global interest in artificial intelligence accelerated, investors began searching for companies that could become major players in the next generation of technology.
Zhipu AI attracted significant investment from technology companies and investment funds. Its strong research background, rapid model development, and continued work on open-source AI helped build confidence around the company.
The release of new GLM models and the company’s open-source strategy also generated significant attention. As expectations grew, Zhipu AI became one of the companies viewed as strategically important to China’s AI ecosystem.
For a period, the market treated Zhipu AI as a major AI growth story.
However, high expectations can also create high pressure.
Why Has Zhipu AI’s Stock Fallen So Sharply?
A decline of approximately 67% from a recent peak does not automatically mean that Zhipu AI’s technology has failed or that the company has lost its relevance.
Instead, the decline may reflect a broader shift in how investors are evaluating AI companies.
During the early stages of an AI boom, investors often focus heavily on technological potential, model performance, research quality, and future growth. But as the industry matures, the market begins asking more difficult questions:
Can the company generate sustainable revenue?
Can its AI models attract long-term users and enterprise customers?
Does the company have a clear path toward profitability?
Can it maintain a competitive advantage as new models continue to enter the market?
These questions are becoming increasingly important as competition in China’s AI industry intensifies.
Kimi, Qwen, DeepSeek, and Tencent Hunyuan are continuously introducing more capable models. Some companies are using open-weight strategies to expand adoption, while others are leveraging large ecosystems, cloud infrastructure, enterprise products, or existing consumer platforms.
As a result, technological progress alone may no longer be enough to justify a high valuation.
The AI Race Is Also a Capital Market Race
The current AI competition is not limited to benchmark leaderboards.
A model can achieve impressive benchmark scores and generate enormous attention, but attention does not always translate into sustainable business growth.
To maintain a strong valuation, AI companies must demonstrate more than research excellence. They need to build useful products, attract customers, create reliable revenue streams, and establish a long-term competitive position.
This is where the competition becomes more difficult.
Large technology companies may have advantages because they already possess cloud infrastructure, large user bases, enterprise relationships, and significant financial resources.
Meanwhile, smaller AI labs must prove that their technology can become a scalable and profitable business.
For Zhipu AI, the challenge may not be developing capable models. The company has already demonstrated strong research capabilities through the GLM series.
The larger question is whether those capabilities can be converted into long-term commercial success.
What Could Happen Next?
Despite the recent decline, it may be too early to view Zhipu AI as a declining company.
The AI industry is still developing, and market sentiment can change rapidly. A new model release, major enterprise partnership, successful commercial product, or stronger revenue growth could significantly improve investor confidence.
At the same time, the company will need to compete in an environment where technological leadership can change quickly.
The next phase of the AI race may be defined less by who releases the largest model and more by who creates the most useful products, builds the strongest ecosystem, and develops the most sustainable business model.
Zhipu AI’s journey highlights an important lesson for the entire AI industry:
A strong model can create attention, but long-term value requires commercialization, revenue, and a durable competitive advantage.
The company’s recent stock decline may not represent the end of its AI story. Instead, it could mark a transition from an era driven by expectations to one driven by execution.
As the competition between China’s leading AI labs continues, the market will be watching closely to see whether Zhipu AI can turn its strong academic foundation and advanced GLM models into lasting commercial success.
