Zhipu AI is a prominent large model developer in China’s AI industry. Founded in 2019 as an offshoot of Tsinghua University’s Computer Science Department, the company was built on academic research and a commitment to advancing general artificial intelligence. From its earliest days, Zhipu AI focused on developing large-scale models capable of language processing, reasoning, and multimodal understanding, laying the groundwork for systems designed to support broad cognitive tasks.

The company’s development path reflects a long-term strategy centered on research depth, structured commercialization, and steady model iteration. Rather than treating AI development as a short cycle effort, Zhipu AI views it as a long-distance endeavor that requires persistence, technical rigor, and continuous refinement. This perspective has shaped both its architecture and its market positioning.

Founding Vision  

Zhipu AI grew out of Tsinghua University’s Knowledge Engineering Lab, where engineering transformation and the principle of “Paper to Product” guided applied research. The transition from laboratory work to commercial deployment formed part of the company’s foundation. Its origins reflect a systematic effort to bring academic breakthroughs into practical use within enterprise environments.

The company’s mission centers on building general artificial intelligence systems. This includes cognitive models capable of handling language tasks, reasoning functions, and multimodal information within a unified architecture. From the beginning, the emphasis has remained on scalable foundational systems rather than narrow applications, ensuring continuity across successive generations of models.

GLM Architecture and Model Innovation

Zhipu AI developed its GLM series using an autoregressive blank-filling pretraining framework. This architectural choice provides a consistent technical backbone for model development and supports ongoing improvements across releases. By maintaining structural continuity, the company advances reasoning performance and intelligence capabilities while preserving alignment between research and deployment.

The GLM model family includes versions such as GLM-4.5 and GLM-4.6, which integrate reasoning, encoding, and intelligent agent features into a single system. According to company statements, these models perform strongly on benchmark evaluations and have been applied across practical scenarios in multiple industries. The GLM series has also demonstrated notable strength in coding tasks and has achieved recognition in global open model usage metrics.

Through APIs and platform access, Zhipu AI makes its models available beyond research environments, enabling developers and enterprises to integrate large-model capabilities directly into applications and workflows.

Enterprise Leadership and Market Presence

Zhipu AI has built a broad enterprise footprint. By mid-2025, its GLM model series had served more than 12,000 enterprise clients and 45 million developers worldwide. Nine of China’s top ten internet companies have adopted its services, reflecting strong engagement from leading digital platforms.

In 2024, the company ranked second among China’s general-purpose foundation-model developers by market share and first among independent developers. This positioning highlights both technical credibility and commercial traction within a competitive ecosystem.

The company emphasizes enterprise applications, aligning its development strategy with productivity-focused sectors such as governance, manufacturing, energy, finance, internet services, communications, education, and consumer electronics. This enterprise orientation reflects its belief that foundational AI technologies should land where productive value is concentrated.

IPO Milestone

Zhipu AI reached a major milestone with its listing on the Hong Kong Exchange on Jan. 8, raising HK$4.35 billion and achieving a valuation close to US$9 billion. The listing marked it as China’s first AI software company to go public and signaled a transition into a capital-market-driven phase of development.

The company views public listing as a stage in a longer journey rather than an endpoint. Leadership has described AI development as a marathon requiring sustained investment in technology, infrastructure, and talent. From the outset, preparations for listing were integrated into long-term planning, including commercialization strategies, revenue generation, cost management, and organizational growth.

The IPO highlights the company’s confidence in the industrial scale of large-model technology and its belief that the field has reached a level where public market participation can support continued advancement.

Leadership and Long-Term Commitment

Zhang Peng, founder and CEO, brings more than two decades of academic and engineering experience from Tsinghua University. His background in laboratory research and system deployment reflects the company’s technical roots. Zhang has consistently emphasized stability, controllability, and long-term vision in high-technology development.

Under his leadership, Zhipu AI has pursued a strategy that prioritizes structured growth, disciplined commercialization, and sustained model improvement. The company’s trajectory reflects a belief that artificial intelligence development requires patience, engineering rigor, and alignment between research ambition and enterprise deployment.

Zhipu’s leadership has expressed that achieving general artificial intelligence remains the ultimate goal. The company views its work as part of a broader technological evolution that extends beyond short-term market cycles. By focusing on foundational models, enterprise integration, and systematic architecture, Zhipu AI positions itself as a pioneer in the history of large-model development.

Future Horizons

Zhipu AI continues to advance its GLM architecture while expanding enterprise adoption and model capabilities. With millions of developers and thousands of corporate clients, the company has built a substantial ecosystem around its technology.

Zhipu AI’s trajectory reflects a combination of research depth, commercialization discipline, and strategic timing within China’s AI industry. As one of the earliest major large-model developers to reach the public markets, Zhipu AI stands at the forefront of the transition from laboratory innovation to industrial-scale artificial intelligence deployment.

The company’s stated long-term aspiration remains clear. It seeks to contribute foundational progress toward general artificial intelligence while building a durable enterprise presence. Through sustained research, structured architecture, and capital-market participation, Zhipu AI continues to define its role as a trailblazer in China’s evolving AI era.

Zhang Peng, Founder & CEO, Zhipu AI