The global landscape of education is undergoing a profound transformation as artificial intelligence (AI) technologies penetrate every facet of the sector. In China, a surge of enterprises—ranging from edtech giants to innovative startups—are aggressively deploying AI-driven solutions, with a particular focus on educational large models and intelligent robots. This strategic shift, marked by intensive R&D investments and collaborative partnerships, signals a new era in smart education, where personalized learning and technological integration aim to redefine traditional educational frameworks.
The Surge of Educational Large Models: A Strategic Imperative
The rise of generative AI, epitomized by tools like ChatGPT, has catalyzed a revolution in educational technology. In 2025, the release of China Smart Education White Paper at the World Digital Education Conference underscored the pivotal role of AI in reshaping educational content, teaching methodologies, governance, and formats, designating 2025 as the “元年 of smart education” . Concurrently, a directive from China’s Ministry of Education and eight other departments emphasized the acceleration of AI educational large models, advocating for their application in vertical fields such as ideological and political education, science education, and basic disciplines .
Domestic enterprises have responded with fervor. DouShen Education, for instance, launched “DouShen AI,” an AI-powered educational large model tailored for vertical sectors, featuring modules like private tutor Q&A and writing consultation, in October 2024 . Vision Source (Shiyuan Stock) upgraded its EasiShow teaching large model to version 2.0, trained on over 220 billion tokens of educational data including textbooks, lesson plans, and courseware .
NetEase Youdao has been at the forefront of this movement, leveraging its “ZiYue” large model to drive product 智能化 (intelligentization). In January 2025, the company released “ZiYue-o1,” the first open-source educational reasoning model, solidifying its commitment to long-term innovation. “The evolution of large models is a marathon, not a sprint—expect 持续 (sustained) updates and AI product rollouts in educational verticals,” stated Zhou Feng, CEO of NetEase Youdao . The model has spawned over 10 vertical applications, covering translation, composition review, grammar instruction, oral practice, physical education, and family tutoring, integrated into hardware like AI question-answering pens and dictionary pens .
Cao Peijie, Director of the Digital Education Research Institute at the Chinese Academy of Educational Sciences, highlights the transformative impact: “Large models are shifting education from traditional knowledge transmission to personalized, interactive learning experiences, enabling true 因材施教 (teaching students in accordance with their aptitudes)” .
Intelligent Robots: The New Frontier in Edtech Deployment
Beyond software, the edtech landscape is witnessing a parallel surge in intelligent robot development, with enterprises recognizing the potential of physical AI agents in educational scenarios. Beijing Ruihong Embodied Intelligence Robot Technology recently signed a strategic cooperation agreement with Beijing Yizhuang Experimental Primary School to explore AI robot R&D, curriculum development, and educational applications, aiming to integrate AI, IoT, and educational large models for end-to-end intelligent upgrades in teaching practice and evaluation .
Donggang Co., Ltd. announced upgrades to its “Xiaozhi 1.0” educational robot, enhancing its intelligent interaction modules to better engage students in dynamic learning environments . These initiatives reflect a broader trend: intelligent robots are no longer mere technological novelties but core components of educational ecosystems, designed to facilitate hands-on learning, personalized tutoring, and even special education support.
“Enterprises with hardware backgrounds are betting on educational robots, targeting vocational training and special education—scenarios requiring physical interaction,” explained Zhu Keli, Founding Dean of the Guoyan New Economy Research Institute. “This reflects a business dichotomy: light assets (content-focused large models) versus heavy assets (robot hardware)” . The integration of robots into classrooms marks a pivotal step toward merging virtual AI with tangible educational tools, addressing gaps in hands-on learning and adaptive instruction.
Market Dynamics: Growth Potential and Competitive Landscapes
The “AI+education” market in China is expanding at an unprecedented pace. A report by Changjiang Securities projects the market size to reach 160 billion yuan by 2027 and approach 180 billion yuan by 2030, underscoring its status as a high-growth sector . During the 2025 “618” shopping festival, intelligent learning machines emerged as the fastest-growing category on the What to Buy platform, with GMV surging by over 10 times year-on-year, evidence of burgeoning consumer demand for AI-empowered educational tools .
Competition in this arena is fierce and multifaceted. Tech giants dominate the top tier with the cloud and algorithmic advantages, while listed educational firms leverage course content and user data in the mid-tier, and startups focus on niche scenarios at the base . For example, iFlytek’s smart education products have been adopted in 32 provincial-level administrative regions, serving over 50,000 schools and 130 million teachers and students. Its Spark Teacher Assistant, used by 150,000 teachers in 2,000 schools, has improved teaching design efficiency by 56.52% and courseware production efficiency by 64.18%, with a 93% satisfaction rate .
However, the market lacks unified standards for product quality, teaching effectiveness evaluation, and data labeling, leading to uneven product quality . “The competitive landscape is fluid; innovative enterprises can always displace predecessors with superior applications,” noted Pan Helin, a member of the Expert Committee on Information and Communication Economics at the Ministry of Industry and Information Technology .
Challenges and the Path to Blue Ocean Markets
Despite its promise, the “AI+education” sector faces significant hurdles. The most pressing issue is the mismatch between technical supply and educational scenarios, characterized by gaps in product design, data application, and industrial collaboration . Zhu Keli emphasizes that breaking through requires three core capabilities: scene coupling, human-machine collaboration design, and effectiveness verification mechanisms .
Another critical challenge lies in balancing technological innovation with educational authenticity. While large models and intelligent robots can enhance accessibility and personalization, there is a risk of prioritizing technology over pedagogical principles. Wu Qi, a senior researcher at the Pangu Think Tank, stresses that competitive edge hinges on application and service capabilities: “Enterprises must focus on core educational problems and user needs, investing in AI technology and educational content R&D to enhance product competitiveness” .
The development of intelligent robots, in particular, presents unique challenges—from high R&D costs to the need for specialized teacher training to integrate robots into curricula. Ensuring these technologies complement rather than replace human educators will be key to their long-term viability.
The Future: Integrating Large Models and Intelligent Robots for Holistic Education
As the “AI+education” ecosystem evolves, the integration of large models and intelligent robots emerges as a strategic priority. Enterprises are increasingly recognizing that these technologies are not mutually exclusive but synergistic—large models provide the cognitive backbone for personalized learning, while intelligent robots offer physical interaction and hands-on engagement.
NetEase Youdao’s approach exemplifies this synergy, with its “ZiYue” model powering software applications and hardware devices alike, including AI-enabled writing tools and robotic tutors. iFlytek’s end-to-end solutions, spanning school teaching, exam systems, and autonomous learning, demonstrate how large models and robotic technologies can work in tandem to create seamless educational experiences .
Looking ahead, the education sector is moving toward a hybrid model where AI technologies—from large language models to intelligent robots—serve as augmentations to human instruction, not replacements. The focus will shift from technological gimmicks to tangible educational outcomes: improved learning efficiency, personalized skill development, and equitable access to quality education.
Conclusion: A New Era of Intelligent Education
The convergence of large models and intelligent robots in education represents more than a technological trend; it signals a fundamental reimagining of how knowledge is created, delivered, and absorbed. As Chinese enterprises continue to invest in “AI+education,” the sector stands at the cusp of transforming traditional classrooms into intelligent, adaptive learning environments.
The road ahead is not without challenges—standardization, scenario adaptation, and ethical considerations must be addressed. Yet, with strategic innovation and a focus on educational value, the integration of AI technologies, particularly intelligent robots, holds the promise of unlocking education’s full potential: to educate, empower, and equip learners for the demands of the digital age. In this new era, the interplay between human expertise and artificial intelligence will define the future of education—one where intelligent robots serve as partners in learning, enabling tailored education at scale and driving educational equity forward.