Chinese AI companies just completed their largest funding bonanza in years, with DeepSeek, Moonshot AI, and StepFun collectively raising $11.5 billion in a matter of weeks. This funding surge marks a dramatic reversal for China's tech sector, which has faced years of regulatory headwinds and venture capital drought. The deals signal that investors now believe Chinese AI companies have cracked the code on commercializing artificial intelligence in the age of AI agents.

Why Are Chinese AI Companies Suddenly So Valuable?

The timing of these funding rounds reveals a critical shift in how the AI industry is evolving. Chinese companies like DeepSeek, Moonshot, and StepFun have moved beyond simply building large language models (LLMs), which are AI systems trained on vast amounts of text data. They're now focused on creating AI agents, software that can autonomously complete complex tasks like ordering food, booking hotels, or managing logistics. This practical focus is attracting billions in investor capital.

DeepSeek's funding story is particularly striking. Founded in mid-2023 and backed by HighFlyer, a Chinese quantitative trading firm managing over $10 billion in assets, the company had never raised external funding before. Yet in April 2025, it launched its first-ever funding round, initially seeking $300 million at a $10 billion valuation. Investor demand was so intense that the round ballooned to $7 billion at a $50 billion valuation, making it the largest single funding round for a Chinese AI company in history.

What changed? DeepSeek faced a talent retention crisis. As the company's engineers and researchers gained prominence after releasing R1, a reasoning model that made global headlines in early 2025, competitors began aggressively recruiting them. Without external funding and formal equity incentives, DeepSeek couldn't compete for talent. The company also needed capital to scale its latest model, V4, which is nearly twice the size of its predecessor and requires proportionally more data and computing resources.

How Are Chinese AI Companies Winning Market Share?

Moonshot AI and StepFun are pursuing different but equally compelling strategies. Moonshot, developer of the Kimi model family, raised $2 billion at a $20 billion valuation, led by Meituan, the Chinese delivery and services giant. This is Moonshot's third funding round in six months, bringing its total capital raised to $3.9 billion. The company's annual recurring revenue (ARR), mostly from API access to its models, reached $200 million in April 2025.

Moonshot's latest model, Kimi K2.6, is a trillion-parameter model with exceptional coding capabilities and the ability to manage up to 300 sub-agents simultaneously. For context, a trillion parameters means the model contains one trillion individual data points that help it understand and generate language. This makes it one of the best open-weight models available, meaning the model weights are publicly available for developers to use and modify.

StepFun, founded in April 2023 by former Microsoft corporate vice president Jiang Daxin, is pursuing a multimodal strategy. The company is raising $2.5 billion and has shipped 16 different models covering text-to-video generation, voice interaction, image understanding, and multimodal reasoning. Its flagship Step-3 model contains 321 billion total parameters with architectural innovations that reduce computing costs to roughly 22 percent of DeepSeek V3's per-token cost.

What's Driving the AI Arms Race Among China's Tech Giants?

The funding rounds reveal an escalating competition among China's largest technology companies. Alibaba owns approximately 36 percent of Moonshot after investing in early 2024, yet Meituan, Alibaba's fierce rival in food delivery and services, is now leading Moonshot's latest funding round. This signals that AI agents are becoming the primary interface through which young Chinese consumers manage daily life, and no major tech company wants to be left behind.

Meituan is also building its own AI models in-house. Its Longcat 2.0 model was trained on 50,000 to 60,000 Huawei Ascend chips, representing the largest training run on domestic Chinese computing infrastructure to date. This investment reflects Meituan's determination to control its AI destiny rather than rely entirely on third-party models.

StepFun's commercial strategy reveals another dimension of this competition. The company recently appointed Yin Qi, former co-founder and CEO of Megvii, as its new chairman. Yin also chairs Qianli Technology, the autonomous driving subsidiary of Geely Auto. This connection suggests StepFun is positioning itself as a critical technology partner for autonomous vehicles and other advanced applications.

How Do These Valuations Compare to Western AI Companies?

While the Chinese funding rounds are historic for the region, they remain dwarfed by Western AI companies. OpenAI closed a record $122 billion funding round at an $852 billion valuation, while Anthropic raised $30 billion at a $380 billion valuation and is reportedly in talks for a new round at a $900 billion valuation. However, the trajectory matters more than the absolute numbers. Chinese companies are now demonstrating clear paths to revenue and practical applications, which is attracting serious capital.

For comparison, Cursor, an AI-powered code editor, has annual recurring revenue exceeding $2 billion as of March 2026 and is in talks to raise $2 billion at a $50 billion valuation. Moonshot's $200 million in annual recurring revenue at a $20 billion valuation suggests the company has room to grow, especially as its models power more consumer and enterprise applications.

Steps to Understanding China's Open-Weight AI Strategy

• Open-Weight Models: DeepSeek, Moonshot, and StepFun all release open-weight versions of their models, meaning developers worldwide can download and modify them. This contrasts with closed models like OpenAI's GPT-4, which are only accessible through paid APIs. Open-weight models accelerate innovation and allow developers to build custom applications without relying on proprietary platforms.
• Multimodal Capabilities: Chinese companies are investing heavily in models that handle multiple types of data, including text, images, video, and voice. StepFun's 16 multimodal models demonstrate this focus, enabling applications far beyond text-based chat. This breadth of capability makes these models more useful for real-world business applications.
• Agent-Centric Design: Rather than building chatbots that respond to user queries, Chinese AI companies are designing agents that can autonomously complete multi-step tasks. Moonshot's K2.6 model can manage up to 300 sub-agents simultaneously, enabling complex workflows in logistics, customer service, and e-commerce without human intervention.
• Domestic Computing Infrastructure: Companies like Meituan are training models on Chinese-made chips from Huawei rather than relying entirely on US-manufactured GPUs. This reduces dependence on foreign technology and aligns with China's broader push for technological self-sufficiency.

The funding surge also reflects a broader recovery in China's venture capital market. From 2021 to 2025, China's tech sector faced prolonged contraction due to regulatory crackdowns, US-China tensions, and rising exit difficulties. The last time a Chinese private company raised $2 billion or more in a single round was Shein in May 2023, at a reduced valuation of $64 billion, down one-third from its prior round. The recent AI funding rounds suggest that investors now see a clear path forward for Chinese technology companies.

These developments have significant implications for the global AI landscape. Chinese open-weight models are already being used by international developers. Cursor's models are trained in part on top of Moonshot's Kimi K series, demonstrating how Chinese AI advances are influencing global AI development. As Chinese companies continue to raise capital and improve their models, they're likely to capture increasing market share in both domestic and international markets, particularly in Asia.