Deepexi Technology's first semi-annual report: The revenue of its AI business increased by 209%, and the company has achieved profitability in the second quarter.

The rapid penetration of AI in the enterprise segment is being continuously verified.

On July 30, DeepData, an enterprise-level large model AI application solution provider, released its first interim performance report after going public. In the first half of 2026, the company achieved operating revenue of 284 million yuan, a year-on-year increase of 115.0%.

By business segment, "DeepexiOS AI-grade enterprise operating system platform solution", which is directly related to AI applications, has become the core growth engine. This business generated revenue of 226 million yuan in the first half of the year, a year-on-year increase of 209.2%, and its revenue proportion rose from 55.3% in the same period of the previous year to 79.6%. In contrast, the revenue of "FastData enterprise-level data intelligent solution" was 58.038 million yuan, basically flat year-on-year. The company's revenue structure is further tilted towards enterprise-level AI applications.

The improvement on the profit side is also noteworthy. The company disclosed that it achieved a net profit of about 30 million yuan in the second quarter of 2026. In the first half of the year, the company's net loss narrowed to 32 million yuan, a decrease of 89.6%; the adjusted net loss narrowed by 48.5% to 27 million yuan from 52 million yuan.

The improvement in profitability mainly comes from two aspects.

First, revenue expansion and business structure optimization drive gross profit growth. In the first half of 2026, the company's overall gross margin was 56.5%, up 1.5 percentage points year-on-year; among them, the gross margin of DeepexiOS AI-grade enterprise operating system platform solution reached 57.2%, up 2 percentage points year-on-year. Driven by the simultaneous increase in revenue and gross margin, the company's gross profit in the first half of the year reached 160 million yuan, a year-on-year increase of 120.5%.

Second, the scale effect has begun to emerge. In the first half of the year, sales and marketing expenses and R&D expenses increased by 33.2% and 23.8% year-on-year respectively, significantly lower than the revenue growth rate. As a result, the sales expense ratio dropped from about 37.3% to 23.1%. In the second quarter, as revenue increased by more than three times quarter-on-quarter, relevant expenses were more fully amortized, and operating leverage was thus accelerated to release.

It should be noted that compared with the adjusted net loss, the significant narrowing of the net loss under the traditional accounting standard in the financial statements is also affected by factors such as reduced share-based payment, fair value changes in the same period of the previous year and no further occurrence of listing expenses. Therefore, compared with the book profit, the continuous improvement of adjusted loss, gross margin and expense ratio can better reflect the profitability trend of the main business.

In general, the single-quarter profitability in the second quarter indicates that DeepData's revenue scale has begun to cross the phased break-even point. If order delivery and revenue growth continue in the second half of the year, and gross margin and expense ratio remain stable, the possibility of the company turning a profit for the full year of 2026 will increase significantly.