AI investment & startup in China Brief — 2026-09-24
Top Stories
1. DeepSeek’s annualized revenue run rate reaches $1 billion as it prepares a major funding round
- Source: Reuters · 2026-09-24
- Summary: Chinese AI startup DeepSeek’s annualized revenue run rate has reached $1 billion, more than doubling within months, according to information cited by Reuters. The company is reportedly targeting a second funding round of roughly RMB 50 billion at a RMB 500 billion valuation, while preparing for further expansion and a potential public listing. The reported revenue acceleration follows significant increases in model pricing.
- Why It Matters: The development indicates that China’s leading model companies are increasingly moving from capital-intensive model development toward substantial commercial monetization. A potential RMB 50 billion financing would also reinforce the concentration of AI capital around a small number of Chinese frontier-model companies.
- URL: https://www.reuters.com/world/asia-pacific/chinas-deepseek-annualised-revenue-hits-1-billion-information-reports-2026-09-24/
2. ByteDance-spun-out AI drug discovery startup Anew Labs raises $290 million
- Source: Jiemian · 2026-09-24
- Summary: Anew Labs, the AI drug-discovery company spun out of ByteDance’s internal AI-for-Science operation, has completed a $290 million first external financing round at a reported post-money valuation of approximately $1.5 billion. ByteDance retains a 56% stake and continues to provide support, including computing resources. The company combines AI research with pharmaceutical expertise and is building operations across China, the US and Singapore.
- Why It Matters: The deal highlights a growing model for Chinese technology giants to turn internally developed AI capabilities into independently financed deep-tech companies. AI drug discovery is attracting capital not simply as an AI application layer, but as a vertically integrated R&D platform combining models, proprietary data and wet-lab capabilities.
- URL: https://www.jiemian.com/article/15136732.html
3. Hangzhou world-model startup Moxin Technology reportedly approaches RMB 10 billion valuation
- Source: 36Kr · 2026-09-24
- Summary: Hangzhou-based 4D world-model startup Moxin Technology is reportedly close to completing a new financing round of around RMB 1 billion, which could push its overall valuation toward RMB 10 billion. The company reportedly completed five financing rounds within nine months, with its valuation rising from about RMB 840 million in March to nearly RMB 10 billion. Its founder Chen Tianrun is a Zhejiang University doctoral student working on 4D world-model technology.
- Why It Matters: The rapid valuation expansion illustrates how aggressively capital is pricing scarce technical talent and emerging world-model capabilities in China. It also shows that investors are increasingly funding models designed to understand and generate dynamic physical environments, rather than focusing solely on language models.
- URL: https://eu.36kr.com/zh/p/3997156704587656
4. Simate raises hundreds of millions of yuan while pursuing AI-driven Physical AI research
- Source: 36Kr · 2026-09-24
- Summary: Physical AI startup Simate has accumulated financing of hundreds of millions of yuan only months after its establishment. The company says its general-purpose physical intelligence model, developed partly with an AI-based automated research system, has reached the top of the RoboDojo benchmark and demonstrated long-horizon tasks, memory and precision manipulation on real robots. Its founding team combines experience in autonomous driving, world models and humanoid-robot control.
- Why It Matters: Simate represents a new investment thesis within China’s Physical AI market: using AI itself to accelerate the research and engineering of robot intelligence. Capital is therefore moving beyond robot hardware toward automated model development, physical data generation and the infrastructure needed to scale robot intelligence.
- URL: https://eu.36kr.com/zh/p/3997018465472388
5. China’s AI capital market is becoming more concentrated around commercially validated companies
- Source: The Paper · 2026-09-24
- Summary: An analysis published by The Paper describes a sharp polarization in China’s AI financing market. While aggregate AI investment remains strong, capital is increasingly concentrating in leading companies and areas with clearer commercial validation, while less-proven startups face greater fundraising pressure. The report highlights the growing investor focus on hard technology and AI applications tied directly to real-world industries.
- Why It Matters: The shift suggests that China’s AI funding cycle is moving from broad model experimentation toward a more selective phase centered on commercialization, technical defensibility and industrial deployment. For startups, access to capital increasingly depends on demonstrating measurable product-market fit rather than AI capability alone.
- URL: https://m.thepaper.cn/newsDetail_forward_34139548.html
6. Physical AI continues to attract unusually rapid financing despite commercialization questions
- Source: Securities Times · 2026-09-24
- Summary: A Securities Times report examines the unusually fast financing cadence in China’s embodied-intelligence sector, with some startups reportedly completing successive funding rounds only one or two months apart. Industry participants also point to a widening gap between financing activity and large-scale commercial deployment. The discussion reflects growing scrutiny of valuations and the ability of robotics companies to convert technical demonstrations into repeatable revenue.
- Why It Matters: The next stage of China’s Physical AI investment cycle is likely to be judged increasingly by deployment economics, production scale and recurring orders. High funding velocity can accelerate technical development, but commercialization milestones are becoming more important inputs into valuation.
- URL: https://4g.stockstar.com/detail/IG2026092400015067
7. China’s AI investment pipeline expands beyond models into AI-for-Science and industrial infrastructure
- Source: Investment界 · 2026-09-24
- Summary: Investment界’s latest AI financing roundup highlights multiple Chinese deals across semiconductor technology, Physical AI, enterprise AI, AI-for-Science and fusion technology. Examples include several-hundred-million-yuan financings for enterprise AI, embodied intelligence and AI-for-Science companies, alongside continued investment in specialized computing and robotics infrastructure.
- Why It Matters: The financing mix indicates that China’s AI investment ecosystem is broadening from foundation models into the infrastructure and vertical technologies required to commercialize AI. The combination of industrial capital and financial investors is particularly significant for startups whose products require hardware, data, compute or domain-specific deployment.
- URL: https://news.pedaily.cn/202609/569619.shtml
8. Shanghai-backed AI-for-Science program brings 42 early-stage teams into an investor ecosystem
- Source: Investment界 · 2026-09-24
- Summary: A Shanghai AI-for-Science startup program brought together 42 early-stage teams and more than 70 participants for a three-day innovation program. The initiative is backed by Shanghai municipal institutions, Shanghai Jiao Tong University, investment organizations and technology partners, with a focus on accelerating commercialization of AI-for-Science projects.
- Why It Matters: Institutional support is becoming an important component of China’s AI startup financing infrastructure, particularly in capital-intensive areas such as scientific computing, biotechnology and industrial AI. Programs connecting researchers, investors and industrial partners can shorten the path from academic research to venture-backed companies.
- URL: https://news.pedaily.cn/202609/569591.shtml
9. DeepSeek’s reported funding plans underscore the widening financing gap between AI leaders and the long tail
- Source: Reuters · 2026-09-24
- Summary: Reuters reports that DeepSeek is pursuing a second funding round targeting approximately RMB 50 billion at a valuation of RMB 500 billion, alongside rapidly rising annualized revenue. The financing effort comes as the company continues investing heavily in model development, with the majority of its computing capacity reportedly still allocated to training. The scale of the proposed round would place DeepSeek among China’s largest AI financing stories.
- Why It Matters: The combination of large-scale financing and rapidly growing revenue is creating a substantially different capital environment for frontier AI companies than for smaller application startups. It also reinforces the importance of compute access and capital intensity in China’s race to develop increasingly capable models.
- URL: https://www.reuters.com/world/asia-pacific/chinas-deepseek-annualised-revenue-hits-1-billion-information-reports-2026-09-24/
10. China’s startup ecosystem increasingly links AI with physical-world data and robotics
- Source: The Yangtzeer · 2026-09-24
- Summary: New startup activity across China’s Yangtze River Delta is increasingly focused on physical-AI data, robotics and AI systems interacting with real-world environments. One emerging example is Maniformer’s crowdsourced physical-AI data platform, which reported 20,000 users and 13,000 task submissions during a one-month trial. The broader ecosystem is moving toward data-generation platforms that support robot training and real-world AI deployment.
- Why It Matters: High-quality physical-world data is becoming a strategic bottleneck for embodied AI. Startups building data collection, evaluation and deployment infrastructure could therefore capture value alongside the robot and model companies themselves.
- URL: https://yangtzeer.com/news/startups/