Leading Across All Metrics: Xunce Technology Reports Record-High Mid-Year Revenue and Net Profit; AI-to-B Platform Enters a New Cycle of Exponential Growth

HONG KONG, Aug 21, 2026 - (ACN Newswire) -On August 21, Xunce Technology (03317.HK) released its 2026 interim results report: In the first half of the year, the company achieved revenue of 967 million yuan, a significant year-over-year increase of 389%, setting a new record high for the same period; net profit attributable to shareholders was 72.5 million yuan, compared to a net loss attributable to shareholders of 89.4 million yuan in the same period last year, marking the company&rsquo;s first profitable first half of the year; Adjusted net profit was 67 million yuan, compared to an adjusted net loss of 105 million yuan in the same period last year, marking a comprehensive turnaround from loss to profit; the gross profit margin reached 60.1%, remaining stable at a high level; total assets were approximately 3514 million yuan, net assets were approximately 2482 million yuan, and with ample cash on hand, reflecting a significant improvement in operational quality.As a leading provider of AI real-time data infrastructure and analytics services in China, Xunce Technology made breakthrough progress in the first half of the year across business models, product innovation, industry penetration, and international expansion. The company led the industry across all key metrics, officially transitioning from the &ldquo;investment phase&rdquo; to the &ldquo;profit realization phase,&rdquo; while also exploring new avenues for profitability.Four Key Drivers Fuel Strong Revenue Growth and Continued Improvements in Operational EfficiencyIn the first half of 2026, against the backdrop of the 15th Five-Year Plan designating artificial intelligence as a national strategic priority, enterprise-level AI is accelerating its transition from model training to practical inference deployment, and from general-purpose capabilities to industry-specific applications. As a result, demand from enterprises for high-quality, structured, scenario-based real-time data is growing exponentially.Facing these significant development opportunities, Xunce Technology based on the FDE model, has expanded its end-to-end data processing and tokenization capabilities&mdash;covering everything from data acquisition to large-model fine-tuning&mdash;to transform dispersed and heterogeneous enterprise data into data tokens. This helps clients reduce costs, improve efficiency, maximize the business value per token, and achieve more agile, data-driven decision-making. Currently, the company has penetrated eleven high-value, high-barrier industries and developed over 400 functional modules. It has refined three core technological advantages: millisecond- and second-level real-time response, 100% accurate data processing, and data tokenization, connecting computing power to algorithms and taking responsibility for clients&rsquo; business outcomes.The company delivered strong performance in the first half of the year, with both revenue and net profit reaching all-time highs. The high revenue growth was primarily driven by the synergistic effect of four key drivers: accelerated deployment of enterprise-grade AI real-time data infrastructure, accelerated penetration into diverse industries, the implementation of the Token business model, and the expansion of international operations and ecosystem development. The significant turnaround in profitability was driven by a steady increase in the proportion of high-margin businesses; the realization of economies of scale from a platform-based and modular product architecture, leading to significantly improved operational efficiency and further optimization of the structure of R&D, sales, and general and administrative expenses; improved cash management efficiency, and investment income contributing to incremental profits.At the same time, the company&rsquo;s operational efficiency has continued to rise. It has established a data development system centered on the AIDP platform, automating data development tasks. Its solutions are deeply integrated into clients&rsquo; self-managed cloud or on-premises systems, serving as a &ldquo;data hub&rdquo; that deeply embeds AI into corporate clients&rsquo; core business processes, supports high-quality business decision-making, and thereby continuously enhances clients&rsquo; willingness to pay;In the first half of the year, the company&rsquo;s ARPU jumped from 1.64 million yuan to 5.56 million yuan, marking a significant year-over-year increase of 240 percent. Customer retention rates have long remained above 90%, fully demonstrating the strong lock-in effect and high customer stickiness resulting from the deep integration of the product into clients&rsquo; business processes, with per-capita revenue surged by 379% year-over-year, placing its labor efficiency among the highest in the AI technology sector.Driven by the dual engines of TokenOS and TokenCloud, revenue from the Token business accounts for over 10%In the first half of this year, the company proactively pioneered the Token business model, enabling data token calls to drive AI-based business decisions. Currently, the Token business model has been implemented in multiple high-value scenarios and has undergone successful commercial validation. During the first half of the year, revenue from the Token business model surpassed 10% of total revenue, with strong growth in ARR (Annual Recurring Revenue). It has become a new engine for revenue growth, and the commercialization of the token economy has far exceeded expectations.The company launched the world&rsquo;s first TokenOS operating system, which focuses on data capabilities. It transforms enterprises&rsquo; multi-source, heterogeneous data in real time into standardized, measurable, and priceable scenario-based tokens, establishing a seamless end-to-end pipeline from data ingestion to model invocation. This enables enterprise data to be directly utilized as tokens and drives the implementation of token factories across various industries.Building on this foundation, the company has also launched the TokenCloud platform, which focuses on model capabilities and &ldquo;data + model&rdquo; capabilities. This accelerates the development of an integrated, end-to-end product and service ecosystem spanning &ldquo;computing power&mdash;data&mdash;tokens&mdash;models&mdash;applications,&rdquo; covering the entire lifecycle from underlying computing power scheduling and data tokenization to model inference optimization, as well as the refinement, tuning, and deployment of enterprise-specific small models. This drives the transformation of AI from a general-purpose capability into enterprise-specific productivity.TokenOS and TokenCloud work in tandem to provide enterprise customers with a one-stop AI infrastructure, enabling them to reduce deployment costs and improve matching efficiency across computing power, data, and models. This &ldquo;operating system + cloud platform&rdquo; dual-drive architecture forms the company&rsquo;s strong competitive barrier, helping it become the computing infrastructure partner for enterprise AI transformation.Accelerated Penetration Across Diverse Industries and Substantial Progress Toward InternationalizationXunce Technology began in the asset management industry, which has the most stringent requirements for data real-time performance. It has now extended its AI real-time data processing capabilities to eleven high-value, high-barrier industries, including telecommunications, electric power, energy, urban operations, and high-end manufacturing. The company is accelerating its penetration across diverse industries, and its ability to replicate solutions across sectors continues to be validated.In the first half of this year, the company collaborated with PATEO Connect and Saimo Technology to jointly develop a token-based physical AI and world model, entering the smart connected vehicle sector; it partnered with three major domestic GPU manufacturers&mdash;MetaX, Iluvatar CoreX, and Biren Technology&mdash;to build a &ldquo;computing power + data&rdquo; closed-loop system; and through a strategic partnership with Shenzhen Kaihong, it implemented tokenization capabilities within the HarmonyOS ecosystem to explore world models and physical AI. The &ldquo;Scenario Token Factory&rdquo; model continues to be validated across multiple industries, scenarios, and ecosystems. As a result, the company&rsquo;s revenue from diversified industries rose to 87.6%, making its business structure more diversified and opening up broader growth opportunities for the company.In addition, the company is actively expanding its overseas operations, accelerating the global rollout of the TokenOS operating system and its tokenization business model; it has also entered into strategic partnerships with the Shenzhen Data Exchange and the Beijing International Data Exchange, laying the foundation for global expansion. In the first half of the year, the company&rsquo;s overseas revenue reached 11.86 million yuan，year-over-year increase of 264%.Four-Stage Evolution of AI Data Infrastructure: Strategic Initiatives Across Five Key DirectionsLooking ahead, Xunce Technology has established a strategic evolution path for the four-stage transformation of AI data infrastructure: starting with data governance (1.0), progressing through the tokenization phase (2.0), and moving toward a global token exchange platform (3.0) in the future, while accelerating the training and large-scale deployment of enterprise-level small models (4.0).Specifically, in Phase 1.0, the company began in the asset management industry and expanded across sectors, fully validating its cross-industry data governance capabilities. In Phase 2.0, the company used TokenOS to refine enterprise private data into high-quality scenario-based tokens and provided one-stop services through the end-to-end integrated cloud service platform TokenCloud, opening up a growth channel for its token business. In Phase 3.0, the company will launch a global Token exchange platform to break down barriers to Token circulation across enterprises, industries, and scenarios. In Phase 4.0, the company will use enterprise-grade small models to bridge the &ldquo;last mile&rdquo; of AI implementation, forming a complete closed-loop ecosystem of &ldquo;computing power&mdash;data&mdash;Tokens&mdash;models&mdash;applications.&rdquo;At the same time, the company will implement strategic initiatives across five key areas: First, deepening the evolution of its business model by accelerating the transition from project-based and subscription-based models to exploring a Token-based business model; second, accelerating cross-industry replication through a &ldquo;high barriers, high value&rdquo; expansion strategy to continuously unlock growth opportunities across diverse industries; third, pioneering cutting-edge applications to secure a leading position in future industries; fourth, building a strategic partnership ecosystem to forge deep ties with upstream and downstream players in computing power and algorithms; and fifth, steadily expanding overseas operations to lay the groundwork for a global footprint. By implementing strategic initiatives across these five key areas, the company will continue to create commercial value for its customers and usher in a new cycle of exponential growth.These impressive financial results mark a turning point for Xunce Technology as it officially begins to realize its value. At a time when demand for high-quality industry data is surging due to the inference requirements of large AI models, the company has built a solid competitive moat through its proprietary token system and opened up vast opportunities for revenue growth through &ldquo;in-depth cross-industry development and international expansion.&rdquo; In the future, as the ecosystem of token exchange platforms matures and enterprise-level small-model applications are implemented, this will significantly expand the potential for profit growth, propelling Xunce Technology onto a fast track of rapid development. Copyright 2026 JCN Newswire. All rights reserved. www.jcnnewswire.com