As a leading provider of AI real-time data infrastructure and analysis services in China, Xunce Technology is rapidly solidifying its core position in AI data, driven by industry tailwinds, full-chain technological capabilities and diversified growth engines. Amid the unfolding landscape of the intelligent economy, this ten-year industry stalwart is entering a pivotal window for structural revaluation.

Token Value Restructuring: Making Every Data Access Quantifiable and Monetizable

Founded in 2016, Xunce Technology has built a full-chain technological system spanning data acquisition, cleansing, standardization, real-time computing and large- model optimization over a decade of development. With AI Data Agent at its core, the Company specializes in millisecond-level real-time data processing, serving a diversified portfolio of industries including finance, urban governance, high-end manufacturing, healthcare, robotics, satellite applications, low-altitude economy, electric power, power grids and energy.

As the era of AI inference unfolds, Token is evolving from mere “fuel” to a form of “hard currency”. Maximizing the value of each individual Token has emerged as a central challenge in the large-model inference era. Today, general-purpose large models typically rely on a “computing power for precision” approach, where every inference run generates substantial wasteful Token consumption. Should inference fail, all Tokens expended in the process are lost entirely, which is a common pain-point plaguing general AI systems.

By contrast, vertical AI solutions equip general large language models with an external industry “brain” powered by domain-specific data. At its core, such solutions optimize inference logic via business-aware models, enabling upfront task feasibility validation and eliminating Token waste at the source. With deep expertise in professional vertical domain data modeling, Xunce Technology leverages its extensive portfolio of high-quality, scenario-specific proprietary data to act as an “efficiency multiplier” for every Token invocation. This structure translates Token consumption into higher-precision outputs while securing maximal result certainty. Crucially, the Company is developing full-chain capabilities spanning data metering, pricing and settlement, enabling quantifiable and monetizable measurement for every data access. By elevating per-unit Token efficiency, it delivers enhanced business value to enterprise clients.

Aligned with this strategy, Xunce’s platform features a “LEGO-inspired” modular architecture, enabling clients to flexibly compose modules tailored to their specific needs. This “assemble-on-demand, adapt-in-real-time” design fosters deep and long-term customer stickiness. The Company also employs a highly flexible pricing framework, with fees structured around module count, processing throughput and other key metrics. Supported by subscription, transaction-based and Token-based payment models, its pricing mechanism precisely aligns with diverse client demands.

Currently, Xunce Technology is fully building a full-chain data measurement and settlement system. It is exploring pricing mechanisms tied to large model inference frequency and module usage count, allowing customers to pay for effective Tokens rather than raw computing power consumption.

Inflection Point Reached, Profitability Confirmed

Driven by Token value restructuring and innovative business models, Xunce Technology has delivered robust performance and reached a historic inflection point. In H2 2025, the Company posted an adjusted net profit of RMB 50 million, achieving its first positive profitability. Meanwhile, revenue rose from RMB 197.85 million in H1 2025 to RMB 1,086.81 million in H2, representing a 449.32% quarter-on-quarter surge. Amid rapid business expansion, the Company has witnessed a substantial improvement in profitability.

For the full year, the Company posted total operating revenue of RMB 1,284.66 million, representing a substantial year-on-year increase of 103.28% and successfully breaking the key milestone of RMB 1 billion in revenue. This signifies that the Company has evolved from an early-stage, technology-driven startup into a new era of platform-based development with scalable and replicable business models.

Furthermore, the Company’s combined gross proceeds for 2025 amounted to approximately RMB 792.08 million, representing a substantial increase of 63.44% compared to approximately RMB  484.63 million in the previous year. In terms of adjusted net loss, after deducting one-off non-recurring gains and losses, the Company’s adjusted net loss for 2025 was RMB 54.84 million, representing a significant narrowing of 33% from RMB 82.37 million in 2024.

Notably, the Company achieved combined gross margin of 62% in 2025, exceeding that of Cambricon (55%), a leading AI chip provider, and far outpacing general large- model developer Minimax (25.4%). This underscores its high-value strategic positioning and resilient business model in the AI data infrastructure sector.

In terms of R&D investment, Xunce Technology has also maintained efficient growth conversion. In 2025, the Company’s R&D expenditure reached RMB 450.44 million, with R&D expenses accounting for 48% of revenue, driving a 105% year-on-year increase in operating revenue. For comparison, Minimax’s R&D expenditure accounted for as high as 219% of its revenue, with a revenue growth rate of 159%. Xunce achieved a comparable expansion pace with a lower R&D intensity.

As revenue scale continues to expand and gross margin in new industries gradually stabilize, the Company’s short-term objective is to achieve an inflection point in adjusted net profit. Looking ahead, as the industries already deployed enter a period of margin stabilization and Token-based payment and revenue-sharing models gain accelerated traction, its net profit is poised to improve at an accelerated rate.

Driven by Diversified Growth Engines, Business Structure Continuously Optimized

The robust revenue growth is not accidental, but underpinned by Xunce Technology’s multi-dimensional, systematically structured growth framework.

Accelerate cross-industry replication. The Company currently covers 9 major industries, benchmarking Palantir’s 17 industries and leaving ample room for horizontal expansion. Xunce Technology is rapidly deepening its presence in key national strategic sectors such as asset management, telecommunications, electric power, urban management, high-end manufacturing, healthcare, energy, robotics training platforms and commercial aerospace. For each new industry, the Company first completes industry-specific data accumulation over 3 to 5 years, enabling rapid replication and scaled deployment across peer customers thereafter.

The business model fosters deep customer value enhancement. As customers evolve from single-module adoption to multi-module deployment, and from pilot trials to full integration into core business workflows, substantial upside potential in ARPU remains. By synergistically lifting Token invocation volume, module usage count and per-Token value, the Company will unlock a new dimension of growth.

Steadily expand overseas business and establish a global layout. The Company targets raising its overseas revenue share to 10% to 15% in 2026, and will further escalate its globalization strategy during 2027 and 2028, unlocking new avenues for sustained long-term growth.

Cultivate a strategic cooperative ecosystem to forge deep integration with upstream and downstream partners in computing power and algorithms. Xunce Technology is engaging in in-depth collaboration with leading domestic GPU providers and large language model enterprises to build a one-stop solution encompassing “infrastructure computing power, upper-layer applications and data governance,” further solidifying its core position in the AI data infrastructure sector.      

Pioneer cutting-edge applications to seize commanding heights in future industries. From robotics data platforms to commercial aerospace, low-altitude economy and power grid systems, Xunce Technology has taken the lead in extending AI infrastructure to emerging sectors with stringent demands for real-time performance and operational reliability. Such mission-critical scenarios with ultra-high requirements for data timeliness and stability serve as a rigorous validation of the Company’s technological advantages. The Company will continue to ramp up R&D investment in frontier fields, refine its technical capabilities through high-end application scenarios and unlock new high-growth, high-value growth tracks for long-term development.

Evolving from Data Services to Core AI Economy Infrastructure: Building Sustainable Competitive Barriers

From a macro perspective, the artificial intelligence data sector is witnessing a profound convergence of five defining trends: a surge in demand for real-time, secure, high-quality data in the era of AI Agent; the rise of domain-specific models elevating professional data as a critical enabler for industry-wide intelligent upgrading; standardization of data interfaces driven by next-generation AI operating systems including Open Claw, positioning Xunce Technology as a core data Token provider; Token-based payment emerging as a new paradigm for the data element market; and the implementation of data asset capitalization policies, spurring a sharp rise in enterprises’ mandatory investment in data governance.

At the intersection of these five pivotal trends, Xunce Technology has established a robust fundamental logic for sustained long-term growth. The Company has evolved beyond a traditional data infrastructure provider to become a critical “connector” and “enabler” linking large models, computing power and cloud vendors. By integrating upstream models, downstream computing power and horizontal collaboration with cloud vendors, the Company delivers irreplaceable data-centric value to its enterprise clients.

The Company stresses that it shares a natural upstream-downstream synergy with general large-model providers, rather than a competitive dynamic. Analogous to the deep collaboration between GPU manufacturers and model developers, the value of Xunce Technology lies in a mutually reinforcing cycle: the wider the adoption of models by its clients, the greater the opportunity for the Company to deliver services and generate incremental value for clients.

Notably, in contrast to niche market players offering only isolated modules such as data cleansing or computing engines, Xunce Technology’s core differentiation lies in its full-process coverage and outcome-driven accountability. The Company provides an end-to-end solution spanning data acquisition, cleansing, standardization, modeling, real-time computing and model tuning, ensuring that final data delivered to clients is clean, accurate, real-time and accessible for model invocation at millisecond latency. Furthermore, through deep integration into clients’ private clouds or on-premises environments, the Company acts as a dedicated data steward, fostering exceptional customer stickiness and forging robust, sustainable competitive barriers.

Currently, the Company’s product portfolio and solutions feature more than 300 functional modules, covering a full spectrum of scenarios from data infrastructure to upper-layer analytics applications. In 2025, its active paid clients base reached 230, with an outstanding customer retention rate of 90%. ARPU increased substantially from RMB 2.72 million in 2024 to RMB 5.59 million in 2025, representing a year-on-year growth of over 103%.

As algorithms become increasingly open-source and computing power tends toward standardization, what truly sets companies apart lies in data — particularly industry data that has undergone sophisticated governance and can effectively empower large models. Backed by a decade of deep industry expertise, Xunce Technology has established a substantial and sustainable competitive barrier in this domain.

Conclusion

From an early-stage private equity tool provider to a cross-industry AI data infrastructure builder, and from a module supplier to a Token-priced core platform, Xunce Technology has remained committed to unlocking data as a scarce, strategic resource that is freely circulable, readily callable and empowered to drive high-quality decision-making.

Amid the rapid emergence of new intelligent economic paradigms, the Company stands at the threshold of a fundamental structural revaluation, with the potential to enter the RMB 100-billion market cap camp. It is a competitor to none, but an indispensable partner to all.