SenseTime's 2025 revenue exceeds 5 billion: Half-year EBITDA turns positive, Generative AI revenue increases by 51%

Ask AI · What are the key drivers behind SenseTime’s EBITDA turning positive?

Leidi Network, Lei Jianping, March 24

SenseTime Technology (0020.HK) today released its annual performance announcement. The financial report shows that SenseTime’s revenue for 2025 is 5.014 billion yuan, an increase of 32.9% from 3.772 billion yuan in the same period last year, marking the fastest growth in three years.

SenseTime’s gross profit for 2025 is 2.056 billion yuan, an increase of 27% from 1.62 billion yuan in the same period last year.

SenseTime’s annual loss for 2025 narrowed by 58.6%. The adjusted EBITDA for 2025 is -645 million yuan, an improvement of 79.3% from -3.117 billion yuan in the same period last year.

SenseTime’s adjusted net loss has seen a continuous acceleration in reduction for four consecutive half-years. The EBITDA for the second half of 2025 is 380 million yuan, marking the first positive result since going public. The company is transitioning from a technology investment phase to a sustainable explosive growth phase.

During the reporting period, SenseTime’s self-sufficiency has significantly improved, with trade receivables reaching 4.87 billion yuan, and the cash conversion cycle (CCC) effectively improving from 228 days at the end of 2024 to 129 days by the end of 2025. In the second half of 2025, SenseTime achieved a positive net cash inflow from operating activities for the first time since its IPO.

Xu Li, Chairman and CEO of SenseTime, stated: “We firmly believe that the deep integration of language and vision is an efficient path to breaking through the limits of intelligence. Based on the NEO native architecture, we efficiently achieve a unified understanding and generation, exploring a new ‘law of scales’ in multi-modal applications. This series of technological breakthroughs, combined with deep AI agents, will unlock new application possibilities, empowering new vertical scenarios. While maintaining innovation in underlying technologies, SenseTime has achieved dual growth in revenue and EBITDA, demonstrating exceptional growth resilience and operational efficiency in the industry, and steadily progressing towards high-quality development.”

SenseTime’s 2025 revenue from generative AI reaches 3.6 billion, up 51% year-on-year

SenseTime’s revenue is primarily divided into generative AI, visual AI, and X innovation business.

SenseTime’s revenue from generative AI in 2025 is 3.63 billion yuan, an increase of 51% from 2.4 billion yuan in the same period last year, driven by explosive growth in market demand for generative AI model training, fine-tuning, and inference. Revenue from generative AI accounts for 72.4% of total revenue, up from 63.7% in the same period last year.

SenseTime’s revenue from visual AI in 2025 is 1.083 billion yuan, an increase of 3.4% from 1.047 billion yuan in the same period last year, benefiting from a domestic demand rebound and continuous growth in overseas markets. Visual AI is entering a secondary growth phase through multi-modal visual intelligence. The revenue growth rate for the second half of 2025 reached 20.9%.

SenseTime’s revenue from X innovation business in 2025 is 300 million yuan. The company stated that it is incubating more X innovation businesses, or that existing X innovation businesses are attracting external investors and becoming independent from the consolidated financial statements. Therefore, in the future, the year-on-year comparison significance of this specific revenue will be reduced.

Deepening native multi-modal large models, “increasing efficiency + reducing costs”

In December 2025, SenseTime released and open-sourced a new native multi-modal model architecture called NEO. SenseTime claims that it only requires one-tenth of the training data and computing power of equivalent industry models to achieve top performance.

Based on the ever-evolving multi-modal large model technology, SenseTime explores the closed-loop value of intelligent agents in strategic areas such as office, finance, marketing, and content generation through a comprehensive dimension of task complexity and fault tolerance, covering clients in various sectors like automotive, smart devices, consumer goods, internet, embodied intelligence, finance, education, and healthcare.

In the general office field, the Xiaohuanxiong family has served a cumulative total of 15 million individual users and thousands of enterprise clients, achieving a sevenfold increase in monthly active users and shortening the business analysis cycle by 90%, thereby enhancing office efficiency.

In the intelligent marketing field, SenseTime has comprehensively covered the full e-commerce process and reached a deep strategic cooperation with Zhi De Mai Technology, improving store operation efficiency by 20 times, live broadcast operation efficiency by 6 times, traffic investment efficiency by 5 times, and real-time monitoring efficiency by 3 times.

In the content creation field, the creative AI agent Seko directly produces up to 3 minutes per episode, continuously generating 100 episodes of content, quickly attracting over 300,000 creators, with monthly active users (MAU) reaching 100,000. The “Financial Intelligent Decision-Making Platform” has rapidly expanded revenue in the banking and insurance industries, while these deep vertical scenario practices have contributed stable growth to the group, also feeding back high-value decision data to enhance the deep logical reasoning ability of the “ever-evolving” model.

Leveraging the core advantages of the “ever-evolving” large model in all-modal interaction, low latency, and long memory, SenseTime promotes the deep embedding of multi-modal large models into various terminal hardware such as smartphones, AI glasses, in-vehicle interactions, and embodied intelligence, with clients including Xiaomi, BYD, and Transsion.

In addition to forging a solid, scalable, and replicable business model in the B2B direction, SenseTime has made progress in exploring a new generation of AI-native consumer applications. The Kapi product series has reached a user scale of over ten million.

Large installations: Deep coordination of underlying computing power and model R&D

Large installations continue to accelerate the synergy of the domestic ecosystem, positioning themselves as the “chain master” of domestic ecological development. They have jointly released the “SenseTime Large Installation Computing Mall” with over ten chip manufacturers, including Huawei Ascend, Haiguang, and Cambrian.

SenseTime states that large installations have become core partners for leading research institutions, internet giants, pan-entertainment companies, embodied intelligence firms, and large model unicorns, and have established China’s first overseas domestic computing cluster in Saudi Arabia.

With support from CATL, the AI intelligent system created by large installations can accurately predict electrical loads through large model computing task data, dynamically generating optimal energy scheduling strategies. It is estimated that the system can achieve a 7% reduction in electricity costs and reduce carbon emissions by more than 4,000 tons per 10,000 P per year. As of the date of this performance announcement, the total operating computing power of large installations has increased to 40,400 PetaFLOPS (FP16).

Visual AI is transitioning from a technology investment phase into a phase of scaled harvest. During the reporting period, CV2.0 achieved profitability for the first time, maintaining positive cash flow for two consecutive years, becoming a pillar for driving group revenue growth and cash flow improvement.

Embracing the global AI industry “finals”

SenseTime states that it will firmly implement the “1+X” strategy in 2025, establishing an efficient collaborative system where the “mother ship (group) provides the foundation, and the child ship (ecosystem) competes in the track,” achieving a leap from single business growth to ecological value multiplication. The ecological enterprises incubated by SenseTime Group have made smooth progress in financing in the primary market, gaining recognition from external capital including internet giants, top venture capitalists, and industry funds. The end-side chip and intelligent driving business segments have completed financing and achieved independent operation off the balance sheet.

SenseTime reveals that the group is prepared to welcome the true “finals” of the global artificial intelligence industry in 2026, continuing to deepen the native multi-modal architecture and establish SenseTime’s position in the integration of native multi-modal large models and spatial intelligence; at the same time, it aims to seize the first-mover advantage in the native AI agent market, achieving a dual explosion of traffic scale and commercial value. The group will also fully promote the adaptation of domestic chips, continuously reducing the inference costs of large models. In the visual AI business sector, it will continue to drive large-scale expansion through dual engines both domestically and internationally. The group will also continue to implement the “1+X” strategy, capturing the added value dividends brought about by the explosion of intelligence in vertical industries.

——————————————————

Leidi was founded by journalist Lei Jianping. Please indicate the source when reprinting.