SenseTime-W (00020) has unveiled its interim results for the six months ended June 30, 2026, posting total revenue of RMB2.9109 billion, a year-on-year increase of 23.4%, with profit attributable to equity holders reaching RMB607 million.

Generative AI revenue surged 28.2% year-on-year to RMB2.3272 billion, lifting its share of total group revenue to 79.9%—a steady climb that underscores how the "one model, one Token factory, one agent control system" framework is accelerating the commercialization of scaled model services, Token offerings, and intelligent agent applications.

Visual AI revenue grew 13.9% to RMB496.8 million, recovering to healthy growth after strategic operational adjustments, while continuing to serve as a critical gateway for the company's expansion into new industries and overseas markets.

Overseas business revenue jumped 127.0% year-on-year, far outpacing overall group growth, reflecting the company's ability to leverage mature visual AI capabilities, unified multimodal models, and localized delivery systems to extend system-level AI solutions to broader regions and client segments.

For the first time, the group disclosed its recurring revenue (RR). In the first half of 2026, RR reached RMB1.1447 billion, up 124.4% from RMB510.1 million in the same period of 2025, and accounted for 39.3% of group revenue—a substantial improvement from 21.6% in the prior-year period. RR refers to revenue derived from contracts effective during the reporting period that carry renewal attributes, and its rapid growth signals a shift in client relationships from one-off project deliveries toward ongoing services and deeper workflow integration, enhancing the sustainability and predictability of the company's income stream.

The group achieved gross profit of RMB1.2062 billion in the first half of 2026, up 32.9% year-on-year, with gross margin improving 2.9 percentage points to 41.4%. This performance stems from continuous optimization of business structure and project quality, alongside joint improvements in model and infrastructure efficiency, higher Token production efficiency, product standardization, and reuse of underlying capabilities. As the unified capability system serves more clients, users, and tasks, the company's earlier technology and infrastructure investments are gradually unlocking economies of scale.