DeepSeek Nears $45B Valuation in First Funding Round

DeepSeek, the Chinese AI startup that shook the global tech industry earlier this year with its highly efficient open-source models, is now in talks for its first-ever external funding round at a valuation approaching $45 billion. The round is being led by China's National Integrated Circuit Industry Investment Fund — commonly known as the 'Big Fund' — marking a dramatic escalation in Beijing's strategic commitment to homegrown AI capabilities.

The news, first reported by the Financial Times on May 6, signals a pivotal shift for a company that has until now operated entirely on internal capital from its parent company, High-Flyer Capital Management (幻方量化), a quantitative trading firm based in Hangzhou.

Key Facts at a Glance

• Valuation: DeepSeek is targeting a valuation close to $45 billion (approximately ¥306.7 billion)
• Lead investor: China's National IC Industry Investment Fund ('Big Fund'), a state-backed semiconductor investment vehicle
• First external round: DeepSeek has never previously raised outside capital
• Parent company: High-Flyer Capital Management, a quantitative hedge fund
• Context: DeepSeek's R1 model shocked the AI world in January 2025 by matching frontier model performance at a fraction of the cost
• Strategic signal: The Chinese government is directly backing DeepSeek as a national AI champion

State-Backed Capital Enters the AI Race

The involvement of China's National IC Fund is perhaps the most significant aspect of this deal. The Big Fund was originally established in 2014 to bolster China's domestic semiconductor industry, and it has since deployed tens of billions of dollars into chip design, manufacturing, and packaging companies.

Its pivot toward investing in an AI model company like DeepSeek represents a broadening of China's tech sovereignty strategy. Rather than focusing solely on hardware, Beijing appears to be extending its state-backed investment apparatus to encompass the AI software layer as well.

This move comes amid intensifying U.S. export controls on advanced chips to China. By investing in DeepSeek — a company that has demonstrated remarkable ability to build competitive AI models using less advanced hardware — China's government is effectively doubling down on a workaround strategy. DeepSeek has shown that algorithmic efficiency can partially compensate for hardware restrictions.

From Hedge Fund Side Project to $45 Billion AI Giant

DeepSeek's origin story remains one of the most unusual in the AI industry. The company was founded as an offshoot of High-Flyer Capital, a Hangzhou-based quantitative trading firm that reportedly accumulated a massive GPU cluster for its trading operations.

High-Flyer's founder, Liang Wenfeng, redirected a significant portion of the firm's computing resources toward AI research, creating DeepSeek as a dedicated AI lab. The company operated without external funding, relying entirely on High-Flyer's profits and existing infrastructure.

This self-funded model gave DeepSeek an unusual degree of independence and speed. Without the need to court investors or meet quarterly expectations, the team could focus on pure research and rapid iteration. The result was a series of increasingly impressive open-source models that culminated in the DeepSeek-R1 release in January 2025.

That model sent shockwaves through Silicon Valley. R1 demonstrated reasoning capabilities comparable to OpenAI's o1 model while reportedly costing a fraction of the training budget. The release temporarily wiped hundreds of billions of dollars off the market capitalizations of U.S. tech giants, most notably NVIDIA, which saw its stock drop sharply as investors questioned the assumption that AI progress required ever-larger hardware investments.

Why $45 Billion? Putting the Valuation in Context

A $45 billion valuation would place DeepSeek among the most valuable AI startups in the world, though still behind the top U.S. players. For context:

• OpenAI was last valued at approximately $300 billion
• Anthropic has been valued at around $60 billion
• xAI (Elon Musk's venture) raised at a $50 billion valuation
• DeepSeek at $45 billion would surpass Mistral AI (valued around $6.2 billion) and most other non-U.S. AI labs

Compared to its Western counterparts, DeepSeek's valuation is notable for what it represents on a cost-efficiency basis. OpenAI has raised over $20 billion in external capital. Anthropic has raised roughly $15 billion. DeepSeek, by contrast, has raised zero external dollars until now — meaning it achieved a $45 billion implied valuation essentially on the back of self-funded research and open-source releases.

The valuation also reflects the geopolitical premium attached to being China's most prominent AI champion. With the Chinese government directly investing through the Big Fund, DeepSeek is being positioned not just as a commercial venture but as a strategic national asset.

Open Source as a Competitive Weapon

DeepSeek's approach to open source has been central to its rapid rise. Unlike OpenAI — which started as an open research lab but has progressively closed its models — DeepSeek has released its most capable models with open weights, allowing developers worldwide to download, modify, and deploy them.

This strategy has yielded several advantages:

• Rapid global adoption: DeepSeek's models have been downloaded millions of times from platforms like Hugging Face
• Community-driven improvement: Open weights enable the global research community to build on and improve DeepSeek's work
• Competitive pressure on U.S. labs: By offering frontier-class models for free, DeepSeek has forced Western companies to reconsider their pricing and access strategies
• Talent recruitment: Open research attracts top-tier AI researchers who value scientific openness
• Geopolitical influence: Open-source distribution ensures DeepSeek's technology is embedded in global AI infrastructure, making it harder to isolate

The question now is whether external funding — particularly from a state-backed entity — will alter DeepSeek's open-source philosophy. Some analysts speculate that government involvement could introduce pressure to restrict access to certain capabilities or to prioritize national security applications.

What This Means for the Global AI Industry

DeepSeek's funding round carries significant implications for multiple stakeholders across the AI ecosystem.

For U.S. AI companies, the deal reinforces that the competitive threat from China is not diminishing — it is accelerating. Despite export controls on advanced chips, DeepSeek has proven that innovation in training efficiency, architecture design, and data curation can offset hardware disadvantages. A $45 billion war chest will only amplify this capability.

For investors, the round validates the thesis that AI is becoming a bipolar technology race between the U.S. and China. Capital is flowing into AI champions on both sides of the Pacific, and valuations are being driven as much by strategic importance as by revenue fundamentals.

For developers and enterprises, DeepSeek's continued rise means more options in the open-source model ecosystem. Companies building AI applications can increasingly choose from a diverse set of foundation models, reducing dependency on any single provider.

For policymakers in Washington and Brussels, this deal is a wake-up call. The U.S. chip export control strategy was designed to slow China's AI progress. DeepSeek's success — and now its state-backed capitalization — suggests that the strategy may need significant recalibration.

Looking Ahead: What Comes Next for DeepSeek

With fresh capital potentially on the way, several key questions will shape DeepSeek's trajectory in the coming months.

Scaling compute infrastructure is likely the top priority. Despite its efficiency breakthroughs, DeepSeek still faces hardware constraints due to U.S. export controls on NVIDIA's most advanced GPUs. The Big Fund's involvement could facilitate access to domestically produced chips from companies like Huawei's Ascend division or SMIC-fabricated processors.

Talent acquisition will be another focus. DeepSeek has already attracted some of China's best AI researchers, but a $45 billion valuation gives it the currency — through equity compensation — to compete more aggressively for global talent.

Product commercialization may also accelerate. While DeepSeek has primarily focused on foundational research and open-source releases, the pressure to generate revenue will increase with external investors on the cap table. API services, enterprise solutions, and vertical AI applications are all potential revenue streams.

Finally, the geopolitical dimension cannot be ignored. DeepSeek's elevation to a state-backed national champion could trigger regulatory responses from Western governments. The company's models could face scrutiny or restrictions in certain markets, particularly in sectors like defense, critical infrastructure, and government services.

The AI race between the U.S. and China has entered a new phase. DeepSeek's $45 billion funding round is not just a financial milestone — it is a declaration that China intends to compete at the frontier of artificial intelligence, with the full backing of the state. For the global tech industry, the implications will unfold for years to come.