Zhipu's earnings surge as AI commercialization breaks new ground! What are the opportunities?

On August 31, enterprise AI service provider Minglue Technology (02718.HK) released its first unaudited interim results since listing on the Hong Kong Stock Exchange. In the first half of the year, the company accelerated its strategic transformation toward AI-Native agentic services, achieving double-digit growth in overall revenue and a more than six-fold explosive increase in agentic service revenue. Although high R&D investment led to continued overall losses, the elimination of large non-cash losses from fair value changes in preference shares recorded in the previous year resulted in a significant year-on-year narrowing of the book loss for the period.
Following the earnings disclosure, trading in Minglue Technology was active. On September 1, the stock surged 21.25% to close at HK$54.2, with a market capitalization of HK$8.286 billion.
According to the interim results announcement disclosed by the Hong Kong Stock Exchange, for the six months ended June 30, 2026, the Group achieved revenue of RMB 760 million, a year-on-year increase of 18.0%; gross profit was RMB 390 million, up 8.3% year-on-year, with a comprehensive gross margin of 51.3% (55.9% in the same period last year); the loss for the period was RMB 64 million, compared to a loss of RMB 204 million in the same period last year, representing a 68.5% year-on-year narrowing in losses; non-HKFRS adjusted net loss was RMB 59 million, compared to an adjusted net profit of RMB 25 million in the same period last year; the Board does not recommend the declaration of an interim dividend for the first half of 2026.

The significant narrowing of the book loss this period is primarily due to the large non-cash losses from fair value changes in preference shares, warrants, and convertible notes present in the same period of 2025, which were completely eliminated in the first half of this year.
Mininglamp Technology's business is divided into two major segments: Data Intelligence Services and Agent-based Services. In the first half of the year, revenue from Data Intelligence Services (the traditional core business) amounted to RMB 587 million, a year-on-year decline of 1.9%, accounting for 77.3% of total revenue.
Specifically, revenue from Marketing Intelligence Services reached RMB 344 million, up 4.2% year-on-year, demonstrating resilience amid cautious corporate marketing budgets. Revenue from Operations Intelligence Services stood at RMB 243 million, down 9.4% year-on-year. While its intelligent store operations system covers over 75,000 offline stores, customer procurement has become more cautious due to the divergent recovery in offline consumption.
The Data Intelligence business completed end-to-end AI efficiency enhancements internally. In the first half of the year, comprehensive efficiency improved by 56.1% across 172 delivery scenarios, accumulating a substantial amount of reusable Skill assets.
Revenue from Agent-based Services (the core growth engine) reached RMB 171 million in the first half of the year, surging 605.7% year-on-year. Its share of total revenue rose to 22.5%, making it the strongest driver of the group's growth. The business focuses on social media marketing, adopting a performance-based payment model, with delivery shifting from process-oriented outputs to result-driven outcomes. Notably, the influencer marketing sub-business achieved a customer repurchase rate of 90%, while the average transaction value increased by 2-3 times year-on-year.
R&D expenses totaled RMB 229 million in the first half of the year, a 52.3% year-on-year increase, with key investments directed toward edge-side models, the Octo Agent collaboration network, and iterations of agent-based products. This high level of R&D spending has exerted some pressure on the company's profitability.
As of the end of June 2026, the company held cash and cash equivalents of RMB 997 million, with total liquid funds amounting to RMB 1.143 billion, indicating a relatively ample cash reserve.
Mininglamp Technology is a domestic provider of enterprise-level AI and commercial data intelligence services, currently in a critical transition period from traditional data BI to Agent-based services. While the traditional Data Intelligence core business remains resilient and Agent-based Services have achieved a breakthrough in revenue scale, enterprise AI commercialization is still in its early stages. High R&D expenditures continue to drain cash flow, and the inflection point for scalable profitability in new businesses remains to be verified.
Notably, from a capital market perspective, the company was formally included in the Stock Connect for Hong Kong Stocks in June, further broadening channels for investor participation. Additionally, in late July this year, Mininglamp Technology signed an agreement to acquire a 19.07% equity stake in Pule Software for approximately RMB 859 million, deepening its layout in industry-specific solutions.
Market analysts believe that this acquisition will enable technological and scenario-based complementarity. Leveraging Pule Software's existing customer channels, there is potential to deploy AI-native financial management products, accelerating the commercial implementation of Mininglamp Technology's enterprise AI products and closing the loop from large model R&D to industry delivery.
As a leading hard-tech AI stock listed in Hong Kong, whether Mininglamp Technology-W can win the favor of the judging committee of the Hong Kong Stock 100 Research Center and be shortlisted for the 13th Hong Kong Stock 100 Strong candidates is worth anticipating. It is understood that preparations for the 13th Hong Kong Stock 100 Strong selection are now underway.
Risk Disclaimer: The above content only represents the author's view. It does not represent any position or investment advice of Futu. Futu makes no representation or warranty.Read more
Comments
to post a comment
