SenseTime delivers its strongest performance ever! Revenue exceeds 5 billion yuan

As the artificial intelligence industry gradually shifts from a technological competition to value realization, the competitive logic of enterprises has also transitioned from singular breakthroughs to systemic construction.

In this context, the latest financial report from SenseTime not only shows growth in revenue scale but also a phased improvement in the loss structure of “losing less and less,” indicating a systemic full-stack capability release centered around “computing power - models - applications.”

In 2025, SenseTime achieved total revenue growth of 33%, exceeding 5 billion yuan, reaching a historical high, with the fastest growth rate in nearly three years; the net loss for the year narrowed significantly by 58.6%, with adjusted net losses showing an accelerated reduction for four consecutive half-year periods. In the second half of the year, EBITDA reached 380 million yuan, marking the first positive result since going public, surpassing market expectations.

EBITDA Turns Positive in Half-Year, SenseTime Reaches a Milestone Turning Point

Looking back at 2025, SenseTime delivered a report of significant phased meaning.

The revenue side continued its growth trend, with annual revenue exceeding 5 billion yuan, setting a historical high with a year-on-year increase of 33%, the fastest growth rate in nearly three years. While continuously increasing technology investments, SenseTime completed a key transformation in operational quality: the net loss narrowed significantly by 59% year-on-year, and in the second half of 2025, the company’s EBITDA achieved its first positive result since going public, reaching 380 million yuan.

This combination of “revenue growth + reduced losses” also means that, driven by generative AI, SenseTime is gradually finding a balance between technological investment and commercialization.

More notably, the improvement in operating cash flow saw the group’s operating cash flow achieve positive net inflow for the first time since going public in the second half of 2025. On the cash management front, accounts receivable reached 4.87 billion yuan, a historical high, and the cash conversion cycle (CCC) was significantly reduced from 228 days at the end of 2024 to 129 days at the end of 2025, notably enhancing capital utilization efficiency. As one of SenseTime’s cornerstone businesses, visual AI achieved net profit for the first time during the reporting period and maintained positive cash flow for two consecutive years.

These changes in financial indicators essentially reflect the aggregated potential of the company’s business structure.

In 2025, SenseTime’s various business segments showed varying degrees of improvement.

The generative AI business continued its explosive growth, achieving revenue of 3.6 billion yuan for the year, a substantial year-on-year increase of 51%, positioning it as the group’s core business. After successfully breaking through the boundaries of traditional perceptual intelligence, the visual AI business is gradually shifting from a phase of technological investment to a phase of scaled harvest, becoming a true pillar of revenue growth and cash flow improvement. During the reporting period, the visual AI business achieved revenue of 1.08 billion yuan, a year-on-year increase of 3%, with the growth rate in the second half of 2025 rising to 21%; this indicates that SenseTime is forming a more resilient operating system through mature businesses feeding back into frontier explorations.

The X-innovation business surpassed 300 million yuan in revenue in 2025. During the reporting period, the ecosystem companies incubated by the group gained high recognition in the primary market from internet giants, top venture capital, and industrial funds, with external financing progressing smoothly, demonstrating strong growth resilience.

From the perspective of the capital market, the current valuation system for technology companies is undergoing adjustment. For SenseTime, the rise in revenue and profitability, combined with long-term growth expectations in the AI industry, provides a foundation for attention in a new round of value reassessment. In other words, the improvement in financial data is not only a result of operational achievements but also supports its long-term positioning in the capital market.

SenseTime’s Chairman and CEO Xu Li stated: “We firmly believe that the deep integration of language and vision is an efficient path to breaking the limits of intelligence. Based on the NEO native architecture, we efficiently achieve unity of understanding and generation, exploring a new ‘law of scale’ for multimodality. This series of technological breakthroughs and the deep integration of intelligent agents will open up new application possibilities, empowering entirely new vertical scenarios. While maintaining innovation in underlying technology, SenseTime has also achieved dual growth in revenue and EBITDA, demonstrating outstanding growth resilience and operational efficiency in the industry, steadily moving towards high-quality development.”

The Value of the “1+X” Strategy Emerges, with “Mothership + Subships” Efficiently Collaborating for AI Finals

If financial improvement is the result, then the core logic driving this result comes from the deepening of the “1+X” strategy.

Here, “1” represents the group’s core business, including generative AI and visual AI, forming the company’s technological foundation and revenue base; while “X” points to the innovation businesses incubated by the group that possess independent growth potential, including intelligent driving with cutting-edge world models and production delivery capabilities, high-performance inference GPUs, end-side AI chips enabling the Internet of Everything, the Meta-Radish creating a new category of home AI robots, smart healthcare focused on industry digital transformation, and smart retail, as well as the Xiaoxiao Robot representing breakthroughs in embodied intelligence.

Around the “1” business, SenseTime has further built a three-in-one full-stack capability closed loop of “computing power infrastructure - large model R&D - large model applications.”

In 2025, SenseTime continued to invest in cutting-edge technology R&D, achieving breakthrough progress in large model architecture innovation, training paradigms, inference efficiency, and spatial understanding, maintaining a leading position domestically. The performance of the SenseTime daily new multimodal large model has continued to rise, consistently ranking first, achieving top positions in comprehensive evaluations such as SuperCLUE and OpenCompass in January, May, and December of 2025. Meanwhile, in 2025, SenseTime successively released and open-sourced the SenseNova-SI series of spatial intelligence models, ranking first globally among similar models in several authoritative international spatial intelligence indicators; the open-sourced Kairos-SenseNova became the first to achieve integrated “multimodal understanding - generation - prediction.” In December 2025, SenseTime released and open-sourced a new native multimodal model architecture, NEO, which can achieve top performance with only one-tenth of the training data and computing power of comparable industry models. These breakthrough advancements not only redefine the inference efficiency and training paradigms of models but also mark SenseTime’s formal entry into a new generation of development in the field of multimodal integration.

Based on the leading technology of the daily new multimodal large model, SenseTime explores the closed-loop value of intelligent agents in strategic tracks such as office, finance, marketing, and content generation through the dimensions of task complexity and fault tolerance. Additionally, leveraging the core technical advantages of the “daily new” large model in full-modal interaction, low latency, and long memory, SenseTime promotes the deep embedding of multimodal large models into various terminal hardware, such as smartphones, AI glasses, in-car interaction, and embodied intelligence, with clients including Xiaomi, BYD, and Transsion.

In 2025, the large device achieved a deep leap from technological advantages to an industrial closed loop. Throughout the year, the large device supported nearly one million model R&D tasks, connecting the complete path from underlying hardware to top-level applications, from software stacks to model adaptation. Among them, the LightX2V world model inference system surpassed overseas chip performance on domestic hardware platforms.

The model of “the mothership (group) providing the foundation, and the subships (ecosystem) competing in the track” offers broader development space. By incubating enterprises in different fields, SenseTime extends its technological capabilities across multiple tracks, forming a diversified business structure. This ecological collaboration not only enhances overall risk resilience but also enables the company to capture more growth opportunities in specific segments.

From “1” to “X,” and back from “X” to “1,” SenseTime is forming a self-growing and self-evolving commercial economy, laying a solid foundation for long-term sustainable development.

As the “1+X” strategy enters a value verification period, SenseTime is responding to the macro proposition of “smart economy” with a full-stack layout, deeply coupling financial performance, business structure, and industrial trends, gradually building certainty for the next growth cycle.

Building the Blueprint for the Smart Economy

In 2026, the “smart economy” was officially proposed as a new economic form. Against this macro backdrop, SenseTime’s business layout provides a representative practical sample.

The report states that it will implement new infrastructure projects such as ultra-large-scale intelligent computing clusters and computing-energy collaboration, strengthen national integrated computing power monitoring and scheduling, and support the development of public cloud. This statement elevates computing power from a single technological element to a national strategic infrastructure, and SenseTime’s “large device” is a forward-looking implementation of this strategy.

The large device continues to accelerate domestic ecological collaboration, jointly launching the “SenseTime Large Device Computing Mall” with more than a dozen chip manufacturers including Huawei Ascend, Haiguang, and Cambricon, establishing a complete domestication solution from underlying hardware to top-level applications. It has now become a core partner for leading research institutions, internet giants, pan-entertainment companies, embodied intelligence firms, and large model unicorns, and has landed the first overseas domestic computing power cluster in Saudi Arabia.

At the same time, with the support of CATL, the “computing-energy collaboration” AI intelligent system has achieved a full-chain connection between computing power management and energy scheduling, estimated to bring a 7% saving on electricity costs and over 4,000 tons of carbon reduction per year for each PetaFLOPS. This practice, centered on the “computing power - model - application” closed loop and extending to domestic collaboration and green development, makes the SenseTime large device not only a core competitive advantage for the company but also a public infrastructure empowering the industry and shared across the entire ecosystem.

The report also emphasizes the creation of a new form of the smart economy. It aims to deepen and expand “artificial intelligence +”, promote the accelerated adoption of the new generation of intelligent terminals and intelligent agents, drive large-scale commercialization of artificial intelligence applications in key industries, and cultivate new intelligent native business formats and models.

This is precisely the goal that SenseTime’s “1+X” and “three-in-one” strategies are committed to achieving. At the application level, SenseTime is deeply embedding AI capabilities into specific industry scenarios such as general office, finance, intelligent marketing, and content generation. In diverse scenarios, AI is gradually becoming a key tool for enhancing efficiency and creating value.

In its financial report, SenseTime dissects the strategic priorities for the future. On one hand, it will adhere to the native multimodal large model route, creating an entry point for the native agent era; at the same time, it will optimize computing costs and promote domestication; in the visual AI field, it will further consolidate profitability and establish the “SenseTime Solution” as a global benchmark for the smart industry; in addition, it will continuously release ecological value and realize the dividends of the “1+X” strategy.

As the industry gradually enters deeper waters, a singular technological advantage is no longer sufficient to support long-term growth; true competition will occur between systemic capabilities and commercial efficiency. From this perspective, the financial improvements, strategic implementation, and ecological construction currently demonstrated by SenseTime are forming a new growth paradigm. The core of this paradigm lies not in short-term performance fluctuations but in whether it can continuously establish a positive cycle between technology, business, and industry.

For the entire AI industry, this may also be an important signal: as the smart economy moves from concept to reality, a company’s value is no longer solely determined by technological leadership, but rather by its ability to achieve collaborative evolution among technology, capital, and industry within a complex system. Author: Jin Song

(Edited by: Zhang Yang HN080)

     【Disclaimer】This article only represents the author's personal views and is unrelated to Hexun.com. Hexun.com maintains neutrality regarding the statements and opinions expressed in this article and does not provide any explicit or implicit guarantees regarding the accuracy, reliability, or completeness of the content included. Readers should only use it for reference and assume all responsibilities themselves. Email: news_center@staff.hexun.com