According to Zhitong Finance App News, Yun Zhisheng (09678) announced interim results for the six months ended June 30, 2026, with revenue of 562 million yuan (RMB, same below), up 38.7% year on year; gross profit of 186 million yuan, up 42% year on year; company owners should account for loss of 237 million yuan, a year-on-year decrease of 20.2%.
According to the announcement, the Group's total revenue was 38.7% year-on-year, and the increase in revenue stemmed from a further increase in the ability to commercialize large-scale model technology. Supported by the continuous iterative upgrading of the native smart big model technology, combined with the Group's intelligent service experience in various scenarios, the Group's enterprise intelligent service revenue segment grew steadily, while the big model mobile-token business achieved explosive growth.
Facing a paradigm shift in the industry, the company adheres to the strategy of “strong base model and deep application” and works collaboratively around a three-tier architecture of base capability layer, intelligent construction layer, and value operation layer. At the base capability level, the company released the native smart model U2 during the reporting period. Relying on intelligent computing clusters and post-training driven by high-quality business data, it carried stronger intelligence with less resource consumption, achieved “high intelligence density and high token value”, and settled customer business objects, processes, rules and actions into reusable assets. The intelligent construction layer, the intelligence platform UniAgent OS encapsulates model capabilities as agents with the ability to act, and the business operation platform UniOps abstracts customer business as an operational business map, making the agent's actions evidence-based. At the value operation layer, the company undertakes every delivery in a closed loop with “planning — decision — execution — feedback”. The key steps are decided together with others, and the key points are confirmed by people. The execution results are recovered in real time, experience is accumulated, and continuous optimization is completed to complete the last mile of value realization, bringing measurable revenue growth, cost reduction and efficiency improvement to customers.
Judging from the operating results, the company achieved revenue of RMB 562 million during the reporting period, with a year-on-year increase of 38.7% during the period. Among them, the enterprise intelligence service achieved revenue of RMB 478 million, an increase of 35.7% over the period; the big model call - token business almost started from scratch, achieving revenue of nearly RMB 30 million, with strong growth momentum; and the end-side AI business achieved steady growth, achieving revenue of nearly RMB 53.8 million. Behind the increase in revenue is a simultaneous increase in the level of capacity platformization and asset reuse. The company is following the three dimensions of revenue growth, cost reduction, and efficiency improvement to transform AI capabilities into measurable business value for customers.