DIDA INC (02559) surged over 14% in morning trading. As of press time, the stock was up 14.29% to HK$1.52, with trading volume reaching HK$11.69 million.

On the news front, DIDA INC recently released its interim results for the six months ended June 30, 2025. The group recorded revenue of RMB 286 million; profit attributable to equity shareholders of the company reached RMB 134 million; basic earnings per share were RMB 0.14; and adjusted net profit amounted to RMB 136 million, representing a 4.7% year-on-year increase.

According to the announcement, the overall gross profit margin for the first half reached 67.0%, with the advertising and other services segment performing particularly well, achieving a gross profit margin of 90.1%, up 6.9 percentage points compared to the same period last year. This improvement was primarily attributed to the scaled expansion of programmatic advertising services, which is more cost-effective than traditional in-app advertising placements, driving continuous optimization of monetization efficiency.

Notably, DIDA INC recently announced a series of new developments in AI large model applications. Under the dual-driven approach of continuously upgrading self-built large model capabilities and the new large model aggregation platform, DIDA INC is reconstructing the foundational infrastructure for business development. Currently, the company's large model applications have expanded from customer service to multiple areas including content security, product research and innovation, business development, and intelligent office operations.

Additionally, DIDA INC recently launched the large model aggregation platform "Tianshu System," which integrates over 10 large models including Deepseek, Tongyi Qianwen, Doubao, and Hunyuan, providing generative AI capabilities to assist with daily office work, problem analysis, and code efficiency improvements. Meanwhile, the Tianshu System also supports API integration with AI model processing interfaces to better empower departments that require deep application of large models.