Large AI funding deals occurred in rapid succession across the global AI value chain in the first half of 2026.

According to an incomplete tally by Zhidx, the global AI value chain recorded 165 funding deals worth at least 100 million yuan in the first half of 2026, raising about 2.7 trillion yuan in total. Chinese companies accounted for 77 deals, the highest number worldwide, while US companies recorded 71. By value, US companies raised about 2.46 trillion yuan in total, compared with roughly 126.2 billion yuan for Chinese companies.

Within China, major AI funding deals were concentrated in cities including Beijing, Shanghai, Hangzhou and Shenzhen. Beijing led by deal count with 42 in the first half. Hangzhou had relatively few deals, but ranked first by funding value, driven by DeepSeek’s single 50 billion-yuan round.

Funding continued to flow into foundational capabilities such as foundation models, computing power and inference deployment, while also spreading into applications including video generation, AI agents, healthcare and enterprise services.

By company, DeepSeek, StepFun and Moonshot AI were China’s three largest AI fundraisers in the first half of 2026, raising about 90 billion yuan combined.

The momentum was even stronger under a broader definition of AI startup funding. According to IT Juzi, Chinese AI startups raised more than 300 billion yuan in the first half of 2026. Its tally covered additional fields such as embodied intelligence and chips, as well as deals worth less than 100 million yuan. In just six months, funding for Chinese AI startups surpassed the total for all of 2025.

China Leads the World’s 165 Major AI Deals, While the US Dominates by Value

According to Zhidx, the global AI value chain recorded 165 funding deals worth at least 100 million yuan in the first half of 2026, raising more than 2.7 trillion yuan in total.

China and the US were the world’s two most active AI funding markets by deal count. Chinese companies recorded 77 deals worth at least 100 million yuan, while US companies recorded 71. Together, the two markets accounted for nearly 90% of the total. Companies in the UK, South Korea, Germany, Israel, the Netherlands, France, Argentina, Australia and Saudi Arabia also raised rounds above 100 million yuan, but their deal counts remained far behind those of China and the US.

The US held a much wider lead by funding value. US companies raised about 2.46 trillion yuan through major deals, accounting for roughly 91.2% of the total. Chinese companies raised about 126.3 billion yuan, or 4.7%, while companies in other countries raised about 110 billion yuan, or 4.1%.

The US lead was driven largely by major companies including OpenAI, Anthropic and xAI. OpenAI disclosed two strategic investments totaling $232 billion in the first half, while Anthropic raised a combined $95 billion across two rounds. Before being acquired by SpaceX, xAI completed a $20 billion Series E round in January.

Series A was the most common funding stage, with 31 deals, followed by 22 Series B rounds, 20 seed rounds, 17 strategic investments, 14 angel rounds, and 13 each at Series C and pre-Series A. Overall, funding deals worth at least 100 million yuan were concentrated in companies’ early and middle stages of development.

Beijing Leads China’s 77 Major AI Deals, While Hangzhou Ranks First by Value

China’s AI value chain recorded 77 funding deals worth at least 100 million yuan in the first half of 2026, raising about 126.2 billion yuan in total.

By location, Beijing recorded 42 deals, Shanghai 16, Hangzhou six and Shenzhen five. Guangzhou, Hefei and Suzhou each recorded two, while Jiaxing and Xiamen had one apiece.

Hangzhou, Beijing and Shanghai ranked as the top three cities by funding value. Hangzhou raised about 50.7 billion yuan, driven primarily by DeepSeek’s single 50 billion-yuan round. Beijing raised roughly 46.34 billion yuan and Shanghai about 26.83 billion yuan. Funding totals were relatively smaller in Shenzhen, Hefei, Suzhou, Xiamen, Guangzhou and Jiaxing.

Major domestic AI funding deals remained concentrated at the early and growth stages. The total included 13 angel rounds, 12 pre-Series A rounds, 11 Series A rounds, eight Series B+ rounds and seven seed rounds. Series A+ and Series B each accounted for six deals, strategic investments and Series C for four each, pre-IPO and Series D for two each, and Series E and Series C+ for one each.

The 77 deals involved 58 companies, many of them founded around 2023. DeepSeek, StepFun, Moonshot AI, SiliconFlow and ShengShu Technology are among a new generation of AI startups that grew rapidly after the large-model boom.

DeepSeek Leads as Five Companies Raise Nearly 100 Billion Yuan in Six Months

By funding value, the top five companies in China’s AI value chain in the first half of 2026 were DeepSeek, StepFun, Moonshot AI, Helixon and SiliconFlow.

DeepSeek raised 50 billion yuan during the first half, StepFun about 21.8 billion yuan, Moonshot AI roughly 18.14 billion yuan, Helixon around 5.29 billion yuan, and SiliconFlow more than 3.48 billion yuan.

DeepSeek

DeepSeek was founded in 2023 and is headquartered in Hangzhou. It was established by Liang Wenfeng, founder of quantitative hedge fund High-Flyer. In 2026, the company released a preview of DeepSeek-V4, DeepSeek-V4-Flash, DeepSeek-V4-Pro, a developer preview of DeepSeek Harness, and a multimodal visual-understanding model, continuing to expand its developer ecosystem.

StepFun

StepFun was founded in 2023 and is headquartered in Shanghai. It was established by former Microsoft global vice president Dr. Jiang Daxin, who serves as CEO, while Megvii co-founder Yin Qi serves as chairman.

The company has launched its Step family of general-purpose large models. In May 2026, StepFun released Step 3.7 Flash as open source. In July, it introduced the agent-native Step AOS operating system, the Amoo personal agent and STEPX, a brand for large-model-native AI devices, marking its expansion into AI hardware.

Moonshot AI

Moonshot AI was founded in 2023 and is headquartered in Beijing. Its founder and CEO is Yang Zhilin, who earned his undergraduate degree from Tsinghua University’s Department of Computer Science and his doctorate from Carnegie Mellon University. He previously worked at Google Brain and Meta AI.

In July 2026, Moonshot AI released its flagship Kimi K3 model as open source, further expanding Kimi’s presence in coding and agent-based use cases.

Helixon

Helixon was founded in 2021 and is headquartered in Beijing. Incubated at Tsinghua University’s Institute for AI Industry Research, it develops new drugs using AI. The company uses large models, multimodal data and automated experimentation platforms to advance drug discovery and design.

Founder and CEO Peng Jian is a senior visiting professor at Tsinghua University’s Institute for AI Industry Research. Helixon was also one of the few AI drug-discovery companies among China’s AI funding recipients that raised at least 100 million yuan in the first half.

SiliconFlow

SiliconFlow was founded in 2023 and is headquartered in Beijing. Positioned as a third-party token services provider, it offers API services, dedicated instances, enterprise-grade model-as-a-service platforms and all-in-one large-model appliances.

Founder and CEO Yuan Jinhui previously served as a lead researcher at Microsoft Research Asia. In late June 2026, SiliconFlow filed for a Hong Kong listing, seeking to join the main board under Chapter 18C rules for specialist technology companies.

The five companies raised more than 98.71 billion yuan combined in the first half, accounting for about 78% of China’s major AI funding total. Their businesses span foundation models, AI devices, AI drug discovery and token infrastructure.

AI Applications Claim the Largest Share as Video Generation and Agents Draw Funding

Domestic AI funding is spreading beyond general-purpose large models into applications, infrastructure and new model architectures.

According to Zhidx, AI applications accounted for 33 of China’s funding deals worth at least 100 million yuan in the first half of 2026, the largest share at 43%. World models and physical AI accounted for 15 deals, or 19%; AI infrastructure for 13, or 17%; general-purpose large models for 12, or 16%; and data services for four, or 5%.

Within AI applications, funding was most concentrated in video, image and 3D generation, which recorded 13 deals and accounted for 39% of application funding. AI agents recorded six deals, or 18%, while enterprise applications, AI healthcare and drug development each recorded four, or 12%. AI finance, AI marketing and other fields also attracted funding.

This shows that domestic AI investment is moving beyond a narrow focus on model capabilities toward specific use cases and product deployment. Capital continues to back foundational capabilities including model training, inference deployment, computing power infrastructure and data services, while also seeking commercialization opportunities in video generation, AI agents, enterprise services, healthcare and finance.

Rather than simply showcasing model parameters, the ability to enter specific industries, build products and generate revenue is becoming a key benchmark for AI investors.

Conclusion: Capital Remains Concentrated in Leading Companies and Applications

China and the US accounted for nearly 90% of global AI value-chain funding deals in the first half of 2026 and the overwhelming majority of funding value. The US market was driven by leading model companies such as OpenAI and Anthropic, with larger individual rounds. Chinese AI deals were smaller on average but more numerous, spanning foundation models, video generation, AI infrastructure, AI drug discovery, embodied intelligence and other fields.

Domestic funding remained concentrated in angel, pre-Series A and Series A rounds, with many companies founded after 2023 securing major investments. Capital also shifted markedly toward the application layer: AI applications accounted for more than 40% of domestic deals, with video generation and AI agents emerging as closely watched paths to deployment.

Overall, AI funding is gradually shifting from concept-driven investment toward a dual focus on technical iteration and commercial deployment. The ability to deploy technology, deliver value in specific use cases and demonstrate commercialization potential is becoming increasingly important in how investors assess AI companies.