I. Industry Overview: Divergent Momentum Across the Dual Tracks of AI and New Consumer Sectors
From December 1 to 3, 2025, the global venture capital market exhibited a pattern of “AI leading the rally, consumption diverging.” Data shows that total global financing in the AI sector reached USD 820 million over the past three days, representing a 35% increase month-on-month. Among this, the AI Agent track stood out in particular, with average financing per project exceeding USD 180 million, becoming a core focus of capital allocation. Meanwhile, total financing in the new consumer sector reached USD 510 million, up 12% month-on-month, but capital concentration increased significantly: projects related to smart hardware and digitalized supply chains accounted for as much as 62%, indicating intensifying internal divergence within the sector.
From an overall trend perspective, the surge in AI financing has resonated with rapid technological iteration, with continued policy support further amplifying momentum. Beijing released the Special Subsidy Implementation Guidelines for AI Industrial Deployment, providing subsidies equivalent to 20% of R&D investment for “AI + manufacturing” projects, with a single-project cap of RMB 50 million. Shanghai simultaneously introduced a new tax rebate policy for AI hardware R&D, allowing enterprises purchasing computing equipment to enjoy a 30% VAT refund upon collection, which is expected to drive industry R&D investment growth of 15%–20%. By contrast, in the new consumer sector, policy catalysis has become the key variable. Policies such as Guangzhou’s “home appliance trade-in” subsidies and Hangzhou’s supporting incentives for new energy vehicles have directly driven significant sales growth in high-end white goods and plug-in hybrid vehicles, providing clear expectations for incremental performance among related companies.
II. In-Depth Analysis of the AI Sector:
The Rise of the Agent Track, with Technological Barriers and Scenario Deployment as Valuation Anchors**
1. Financing Cases: Leading Projects Demonstrate Dual Advantages in Technology and Scenarios
Among recent AI financing cases, the explosive growth of Genspark AI is particularly noteworthy. Following the completion of a USD 275 million Series B round, the company reached a post-money valuation of USD 1.25 billion, entering unicorn territory. Notably, it achieved a valuation doubling within just five months of transitioning from AI search to AI Agent services. The core logic lies in Genspark AI’s AI Workspace, which adopts a multi-model collaborative architecture integrating more than 30 mainstream models and 150 proprietary tools. This addresses the pain point of “fragmented usage” inherent in traditional AI tools and enables end-to-end automation from “goal input” to “output delivery.” Its current customer renewal rate stands at 92%, significantly higher than the industry average of 68%. This round was led by Emergence Capital, with global institutions such as SBI Investment participating, and all existing investors increasing their stakes, reflecting strong capital confidence in its technical pathway.
Domestic projects have also demonstrated differentiated competitiveness. Xiyin Intelligence, incubated by the Zhongguancun Two Institutes AI Innovation Fund, completed a RMB 20 million seed round, reaching a post-money valuation of RMB 120 million. Its core product is a digital twin system for medical scenarios, focusing on chronic disease management and rehabilitation monitoring. Leveraging technical collaboration with Tsinghua University and the Chinese Academy of Sciences, it has established a first-mover advantage in vertical scenarios. In addition, Xinghe Technology secured a USD 120 million Series A round led by Sequoia Capital Global Growth Fund, with a post-money valuation of USD 580 million. Its on-device AI inference chips achieve an energy efficiency ratio of 2.8 TOPS/W and show broad application prospects in smart wearables and industrial sensors, positioning the company as an important participant in the on-device AI hardware track.
2. Market Dynamics: Capital Flows Upstream, with Computing Cost Risks Requiring Vigilance
From the perspective of capital flows, the A-share AI hardware sector recorded a net inflow of RMB 8.9 billion over the past three days, with optical modules and on-device chips emerging as the two core directions. The optical module track, represented by companies such as InnoLight and Tianfu Communications, attracted RMB 3.5 billion in inflows, while on-device chip companies such as Allwinner Technology received RMB 2.8 billion. Foreign capital simultaneously increased exposure, with northbound funds net buying RMB 1.27 billion in AI hardware stocks, raising their holding ratio to 5.2%. This trend reflects the accelerated deployment of on-device AI applications and growing overseas demand. For example, InnoLight’s 1.6T optical module products have entered overseas sampling stages, and expectations of mass production in 2026 have driven a wave of brokerage buy recommendations.
However, profitability risks in the AI sector are gradually emerging. Due to an 18% month-on-month increase in NVIDIA H100 chip prices and overseas restrictions on computing power exports, average computing costs for domestic AI enterprises have risen by 22%. Some small and mid-sized AI Agent companies are facing margin pressure due to difficulty absorbing these costs. This raises the bar for future investment screening: enterprises with hybrid-architecture technologies capable of reducing reliance on computing power may gain a more advantageous competitive position.
III. New Consumer Sector:
Hard Technology Empowerment and Policy Catalysts Give Rise to Segment Leaders**
1. Smart Hardware: Technological Breakthroughs Unlock Overseas Market Potential
Within the new consumer sector, smart hardware has become a focal point for capital due to technological barriers and overseas expansion capabilities. Jike Technology recently completed two financing rounds totaling USD 70 million, reaching a post-money valuation close to USD 400 million. Its core product is consumer-grade exoskeletons, with the second-generation product reducing weight to 3.2 kg (industry average: 4.5 kg) and extending battery life to eight hours. The product has successfully entered European and U.S. rehabilitation medical channels, with projected 2026 revenue exceeding USD 500 million. In this round, 40% of funds will be allocated to product iteration to further solidify technological advantages.
La Mou Technology has also delivered strong performance, securing a multi-tens-of-millions RMB Series A+ round led by Jiuyou Capital. Its core product is an intelligent lawn-mowing robot. Leveraging “visual navigation + automatic obstacle avoidance” technology, La Mou has achieved an 8% market share in Europe and the U.S., surpassing Ecovacs at 6%, and emerging as a dark horse in the segment. With a founding team possessing R&D backgrounds at companies such as Narwal, the company demonstrates precise understanding of overseas market demand. This round will fund the construction of localized production facilities in North America, expected to reduce logistics costs by 18% and further enhance product cost-performance.
2. Supply Chain and Services: Digitalization Drives Efficiency Gains
Enterprises related to digitalized supply chains and service upgrades have also attracted capital attention. Ruiyun Cold Chain completed a nearly RMB 100 million Series A+ round. Its digital system has reduced order response times to 15 minutes, far below the industry average of 45 minutes. Its services now cover 100,000 terminal stores, 70% of which are chain catering brands such as Mixue Bingcheng and Guming. Cold-chain loss rates are controlled below 3%, significantly outperforming the industry range of 5%–8%, while order volume has grown 70% year-on-year, demonstrating strong growth momentum.
Tiantian Baiying focuses on intelligent maintenance services for consumer chain brands. After completing a multi-tens-of-millions RMB Series A round, it has helped clients such as Bawang Chaji and Hema reduce equipment failure rates by 30% and lower per-store maintenance costs by 20%. Its services now cover more than 8,000 stores nationwide, with month-on-month growth of 10% for seven consecutive months. Revenue is expected to exceed RMB 200 million in 2025, making it a representative case of digital upgrading in consumer services.
3. Chain Brands: High Revenue per Square Meter Supports Expansion Ambitions
In the chain brand segment, the financing activity of Bixing Coffee has drawn significant attention. The brand secured a multi-tens-of-millions RMB Series B round exclusively led by the Suzhou Agricultural Development Industry Innovation Fund. It currently operates 300 stores concentrated in core commercial districts of East and South China, adopting a “small-store + high revenue density” model. Average monthly revenue per store reaches RMB 250,000, comparable to Manner at RMB 280,000, while gross margins reach 65%, exceeding the industry average of 55%. Following this round, Bixing Coffee plans to open an additional 150 stores to further densify regional market presence and pursue differentiated breakthroughs amid intense competition in the coffee sector.
IV. Investment Research Strategy:
Seizing High-Conviction Opportunities While Guarding Against Potential Risks**
1. Core Allocation Directions
Based on industry trends and corporate competitiveness, it is recommended to prioritize AI Agent enterprises with hybrid-architecture technologies (such as targets along the Genspark AI value chain) and high revenue-density chain brands (such as Bixing Coffee) as core allocations. The former benefit from the high-growth dividend of the AI Agent track, with technological barriers that are difficult to replicate in the short term; the latter rely on refined operational capabilities, offering stronger counter-cyclical resilience amid consumption divergence and higher certainty of earnings growth.
2. Opportunistic Allocation Options
For opportunistic allocations, attention may be given to upstream AI hardware computing equipment suppliers (such as InnoLight) and policy-beneficiary high-end consumer categories. Optical module companies like InnoLight are poised to benefit from the release of overseas demand for 1.6T products, with mass production expectations in 2026 driving earnings growth. Meanwhile, companies such as Midea COLMO (with 80% growth in high-end appliance retail sales) and Anjoy Foods (with 60% growth in prepared food businesses) benefit from policy subsidies while possessing technological or brand barriers, positioning them for dual valuation and earnings uplift during the consumption upgrade cycle.
3. Key Indicators to Monitor Next Month
Looking ahead to late December, two major areas warrant close monitoring. In the AI sector, validation of AI Agent deployments in vertical scenarios such as finance and healthcare, as well as mass-production progress at on-device AI chip companies such as Allwinner Technology and Rockchip, will directly influence valuation logic. In the new consumer sector, sales data for high-end appliances and new energy vehicles during the New Year holiday period, along with the market reception of winter product launches by prepared food companies, will provide key evidence for assessing consumption recovery strength and segment-level prosperity.
In addition, continued vigilance is required regarding the risk of AI computing costs rising beyond expectations, as well as profitability pressure among certain new consumer enterprises caused by overly aggressive expansion. While capturing industry opportunities, refined screening is essential to reduce investment risk.