AI investment & startup in China Brief — 2026-09-25
Today: China’s AI funding ecosystem is broadening beyond model companies, with long-term institutional capital targeting AI and semiconductors while startups push deeper into materials, robotics and commercialization.
Top Stories
1. 📊 China Life plans up to RMB 4.5 billion investment in AI and semiconductor fund
China Securities Journal / Securities Times · September 25, 2026
Bottom line: China Life plans to commit up to RMB 4.5 billion to a new RMB 6 billion fund focused on unlisted AI and semiconductor companies.
China Life and other investors plan to establish the Qingdao Chengxin Zhida Equity Investment Partnership, with total subscribed capital capped at RMB 6 billion. China Life would contribute no more than RMB 4.5 billion, while the fund will target companies with core technology and opportunities in AI and semiconductor supply chains.
Why it matters: The move brings insurance capital further into China’s strategic technology financing ecosystem, potentially providing startups with longer-duration funding beyond traditional venture capital.
2. 🤖 AI-for-materials startup Sogge Intelligent Computing raises tens of millions of yuan
36Kr · September 24, 2026
Bottom line: Shanghai-based Sogge Intelligent Computing has raised a tens-of-millions-yuan angel round to commercialize AI-driven materials discovery.
The company combines an atomic-level materials foundation model, high-performance simulation and wet/dry laboratory experiments to shorten materials R&D cycles. The round was led by Fosun Fortune Capital, with participation from Nantong Industrial Control’s AI fund, Yunqi Partners and Win-Win Venture Capital, alongside existing investors.
Why it matters: AI investment is moving into industrial R&D, where proprietary scientific data, simulation and experimental feedback loops can create defensible application-specific AI systems.
3. 📊 DeepSeek targets a reported $7.5 billion fundraising round as revenue run rate reaches $1 billion
Reuters · September 24, 2026
Bottom line: DeepSeek is reportedly targeting roughly $7.5 billion in new funding at a valuation of about $75 billion while its annualized revenue run rate reaches $1 billion.
The company is said to be finalizing its second funding round and preparing for a potential Shanghai Stock Exchange listing. The reported revenue acceleration followed higher API prices and continued demand for its models.
Why it matters: The combination of substantial recurring AI revenue, large-scale fundraising and IPO preparation signals a shift in China’s frontier-model market toward capital-intensive commercialization rather than purely research-driven competition.
4. 🤖 China’s AI video sector attracts startups and investment as production moves toward commercialization
Reuters · September 25, 2026
Bottom line: Chinese entrepreneurs and investors are increasingly treating AI-generated video as an emerging production industry rather than only a model-development market.
Startups are building AI film studios and production workflows around increasingly capable generative-video models. The activity spans content creation, production services and new business models aimed at turning lower-cost AI generation into commercial output.
Why it matters: The opportunity is shifting toward application-layer companies that can combine models with proprietary workflows, distribution and content economics.
5. 📊 China’s AI investment market remains concentrated among a small group of frontier companies
36Kr · September 24, 2026
Bottom line: DeepSeek’s latest financing highlights the continuing concentration of China’s largest AI funding rounds among frontier-model companies.
36Kr reported that DeepSeek has completed a reported $7.5 billion financing round, alongside annualized revenue exceeding $1 billion. The financing comes as China’s leading AI companies compete for compute, talent and model commercialization opportunities.
Why it matters: Capital concentration gives leading model developers greater capacity to fund compute and research, while pushing smaller startups toward specialized applications, infrastructure and vertical AI.
6. 🤖 Physical AI and robotics continue to attract Chinese startup capital
EarlyTip · September 25, 2026
Bottom line: New Chinese robotics startups are raising early-stage capital around specialized physical-AI applications rather than relying solely on general-purpose humanoid platforms.
Recent deals include seed financing for Xiantuzhiyuan, a robotics company focused on specialized quadruped and mobile manipulation systems, and financing for Xinhe Robotics, which develops intelligent grain-storage robots. Other startups are pursuing task-specific robotics and robot-as-a-service models.
Why it matters: The funding pattern suggests investors are exploring commercially measurable robotics use cases where deployment, data collection and recurring service revenue can reinforce the AI technology stack.
7. 🌐 Chinese AI investment increasingly connects strategic capital with industrial deployment
China Securities Journal / Securities Times · September 25, 2026
Bottom line: China’s latest AI investment activity shows growing alignment between financial capital, industrial policy and commercialization of strategic technologies.
China Life’s planned AI and semiconductor fund is designed to invest in unlisted companies with core technology barriers and opportunities for domestic substitution. The six-year fund structure is explicitly oriented toward technology innovation and strategic industries.
Why it matters: For AI startups, the investor base is becoming broader than conventional venture capital, potentially changing fundraising horizons, ownership structures and commercialization expectations.
8. 📊 China’s AI startup ecosystem is expanding into vertical scientific and industrial applications
36Kr · September 24, 2026
Bottom line: Sogge Intelligent Computing’s financing illustrates how Chinese AI startups are applying foundation-model techniques to specialized scientific workflows.
The company is developing AI systems for materials design and combining computational modeling with laboratory validation. Its investors include both financial and industrial-oriented capital, while the company says it has already worked with leading firms in solid-state batteries and chemicals.
Why it matters: Vertical AI startups can build differentiation through proprietary scientific workflows and domain-specific data rather than competing directly with large general-purpose model providers.
9. 🤖 AI commercialization is becoming a major theme at China’s digital-trade ecosystem
China Securities Journal / Securities Times · September 25, 2026
Bottom line: AI companies at China’s Global Digital Trade Expo are increasingly focused on deploying AI into concrete commercial workflows and overseas markets.
The 2026 expo features more than 2,000 companies, with AI-related businesses accounting for roughly half of exhibitors according to organizers. Reported international procurement orders reached $1.085 billion, with AI-related orders accounting for 53%.
Why it matters: The emphasis on business workflows and cross-border deployment points to a maturing AI startup market where international revenue and enterprise adoption increasingly matter alongside model capability.
10. 📊 China’s AI capital market is widening beyond pure model startups
China Securities Journal / Securities Times · September 25, 2026
Bottom line: New capital commitments and startup activity indicate that China’s AI investment market is spreading across models, industrial AI, robotics, semiconductors and scientific computing.
Recent activity includes large institutional allocations toward AI and semiconductor funds alongside early-stage financing for AI-for-materials and physical-AI startups. This creates a broader financing ecosystem spanning foundational infrastructure and downstream applications.
Why it matters: The next phase of China’s AI startup market is increasingly defined by commercialization, specialized technology and integration with strategic industrial sectors rather than model scale alone.