① The stock price of Lanshiguangbiao, a major AI application company, rose by a cumulative maximum of 107.2% within the month. ② Create a list of A-share companies with high AI application revenue and their percentage (see appendix).
Cailian Press, January 24, by editor Lichen — Due to the rapid advancement of domestic open source large-scale model capabilities, the industry generally defines 2026 as a breakthrough year for the commercialization of AI applications. Yangtze Securities pointed out that AI applications may be accelerating towards realization, and a paradigm shift in models by 2026 may bring extraordinary opportunities. In the long term, there is optimism about the opportunities for upgrading within the AI industry. Guotai Haitong Securities pointed out that applications and computing power will change alternately, and applications will follow the advancement of computing power. This AI industrial revolution is likely to trigger a global synchronized market trend. Towards 2026, AI applications are expected to evolve from highly functional to highly efficient, with the implementation of diverse business models. AI applications are likely to be the core of AI industry trends in 2026. In the secondary market, Lan Shiguangbiao’s stock price increased by up to 107.2% cumulatively during the month.
According to previously compiled Cailian Press VIP Market Tracker data, as of the third quarter of 2025, Lan Shiguangbiao’s AI-driven business revenue reached 2.47 billion yuan, accounting for 4.83%. As of the second quarter of 2025, Zhiyang Innovation’s AI solutions for transmission scenarios reached 381.7 million yuan, accounting for 82.52%. Yunchong Technology’s AI solutions revenue was 127.7 million yuan, accounting for 75.55%. Glingdeep’s AI product revenue reached 24.88 million yuan, accounting for 58.58%. Wanda Software’s AI video big data solution revenue reached 64.72 million yuan, accounting for 50.78%. Toersi’s AI software products and services revenue totaled 108.9 million yuan, accounting for 44.49%. Guanxiang Technology’s AI business revenue was 17.84 million yuan, accounting for 43.45%. Anuoyi’s digital AI, computing power, environmental protection and other industrial businesses earned 163.2 million yuan, accounting for 31.86%. Hanwang Technology’s AI terminal business revenue was 233.4 million yuan, accounting for 28.70%. Zhongke Jinchai’s comprehensive AI service revenue reached 48.35 million yuan, accounting for 14.01%. Chuangye Heima’s AI service division’s AI training business revenue was 9.791 million yuan, accounting for 13.88%. Daotong Technology’s AI and software business revenue totaled 281.1 million yuan, accounting for 11.98%. Sanlian Hongpu’s industrial AI integrated application solutions revenue was 47.41 million yuan, accounting for 9.78%. Hanxin Technology’s industrial AI products and solutions revenue was 4.721 million yuan, accounting for 7.39%. Yuanguang Software’s AI business revenue was 60.88 million yuan, accounting for 5.77%.
As of the fourth quarter of 2024, the revenue of Shensi Electronics’ AI + smart city business reached 142.4 million yuan, accounting for 74.07%. Prologis’ security AI business revenue was 259.4 million yuan, accounting for 35.21%. Nengke Technology’s AI products and services revenue reached 273 million yuan, accounting for 18.08%. Kaipuyun’s AI content security business revenue was 94.94 million yuan, accounting for 15.37%. Chunzhong Technology’s AI product revenue totaled 38.83 million yuan, accounting for 7.85%. Specific details are provided below.
On January 12, Lan Shiguangbiao revealed on its official WeChat account that Lanshiguangbiao’s AI-driven business revenue for the full year 2025 is expected to reach 3 billion to 5 billion yuan (the first three quarters have already reached 2.47 billion yuan). In a research report released on November 16, 2025, Yang Renwen of Kunkai Securities and others stated that Lanse Guangbiao’s AI-based advertising has entered the high-speed development stage. By the end of Q3 2025, the company’s token usage through API integration into large-scale models exceeded 600 billion, and 136 AI agents based on the BlueAI platform were incubated, covering core business scenarios such as video generation, intelligent placement, and content creation to drive cost reduction, efficiency improvement, and revenue growth.
Zhiyang Innovation announced on December 17th in its IR activity record that its embodied intelligence drone products focus on the core requirements of satellite-rejecting space inspection. A multi-stage development was undertaken, including “scenario demand studies – technical route validation – algorithmic breakthroughs – prototype validation – engineering optimization – real world refinements”. The main technical directions and features are: Research on autonomous navigation technology in spaces where satellite reception is not possible. It overcomes the important challenge of high-precision navigation in an environment without satellite signals, and effectively solves the problem of autonomous positioning and continuous operation of inspection equipment in areas where satellite signals cannot be received within a limited space. Research on omnidirectional obstacle avoidance technology. It achieves omnidirectional real-time recognition and intelligent obstacle avoidance in narrow and complex spaces, greatly improving operational safety and area coverage completeness in obstacle-dense environments. Research on precise hazard location in complex environments using advanced AI algorithms. Tackles the challenge of intelligently detecting subtle defects in hidden areas, enabling accurate recognition and quantitative analysis of millimeter-level cracks, corrosion, and other defects, significantly reducing false positive rates.
Yunchong Technology said on the interactive platform on January 16 that the company is steadily promoting the two-wheel drive strategy of “AI infrastructure + AI intelligent agent”, aiming to become an industry growth partner in the intelligent agent era. The company is building foundations and developing products in areas such as healthcare and e-commerce, along with scalable implementations in areas such as smart finance, governance, transportation, and manufacturing, based on the upgraded Composure Knowledge Factory platform CWOS. Successful products include intelligent O&M, intelligent testing, content moderation, intelligent customer service, intelligent health assistants, etc. Technically, the company continues to build an evolvable AI-native technology system around CWOS, building a knowledge factory equally driven by knowledge and data, continually enhancing core capabilities such as model services, information technology adaptation, and intelligent agent collaboration, and laying out an embodied intelligence ecosystem for the future to support scalable replication and continuous evolution across multiple industries. Strategically, the company has transformed its business towards standardized products and subscription services through its “dual curve” strategy, significantly improved its operational quality, and shown an overall solid and positive development trend.
Glingdeep issued an investor relations record on December 2, 2025. In the company’s visual large model Glint-MVT series, Glint-MVTv1.5(RICE) was pre-trained on 1 billion image datasets and 2 billion local regions and outperformed CLIP and SigLIP2 in academic evaluation, and experimental results were presented in a companion paper at the ICCV2025 conference. As a fundamental vision model, Glint-MVT powers downstream tasks such as detection segmentation and OCR, helping product teams in various domains quickly improve the production efficiency and quality of their algorithms. Leveraging robust technology capabilities and extensive commercialization experience, the company’s AI products have been successfully deployed in smart finance, city management, government affairs and specialized sectors, and smart education.
