China's Big Fund Eyes Leading DeepSeek's $45B Round

China's National Integrated Circuit Industry Investment Fund — commonly known as the 'Big Fund' — is reportedly in discussions to lead DeepSeek's first-ever external funding round, a deal that could value the AI startup at approximately $45 billion (HK$351 billion). The move would mark the first time China's most powerful semiconductor investment vehicle has backed a large language model company, signaling a dramatic shift in Beijing's AI investment strategy.

Key Takeaways

• China's Big Fund is in talks to lead DeepSeek's inaugural funding round at a reported $45 billion valuation
• Tencent and other major investors are also negotiating for stakes, though the final investor lineup remains unconfirmed
• DeepSeek founder Liang Wenfeng may personally participate in this funding round
• The company's valuation has more than doubled from $20 billion, reflecting surging investor confidence
• DeepSeek currently lacks a clear commercialization model, yet investors are betting heavily on its technical potential
• The Big Fund's Phase III raised $47 billion, primarily targeting semiconductor equipment and materials

A Historic Pivot for China's Biggest Chip Fund

The Big Fund has traditionally focused exclusively on semiconductor hardware — backing companies like SMIC (Semiconductor Manufacturing International Corporation) and Yangtze Memory Technologies Co. (YMTC) in its mission to reduce China's reliance on foreign chipmakers. A move into AI foundation models would represent a significant departure from this mandate.

The fund launched its Phase III in 2024, raising approximately $47 billion — the largest of its three phases. That capital was earmarked for semiconductor equipment and materials, the critical chokepoints where U.S. export controls have hit China hardest. Until now, the Big Fund has not publicly backed any large language model company.

If the deal closes, it would send an unmistakable signal: Beijing views foundation model development as strategically equivalent to chip manufacturing. This alignment of national semiconductor policy with AI model development could reshape how China allocates resources across its tech ecosystem.

DeepSeek's Valuation Surges Despite No Revenue Model

Perhaps the most striking aspect of this reported deal is the valuation jump from $20 billion to $45 billion — a 125% increase — for a company that has yet to establish a robust commercial business. DeepSeek has made its models freely available and open-source, prioritizing adoption and technical reputation over immediate monetization.

Investors appear undeterred by the lack of revenue. DeepSeek has earned a formidable reputation for its coding capabilities, which are widely considered among the best produced by any Chinese AI lab. Its models have consistently performed well on international benchmarks, competing with offerings from OpenAI, Anthropic, and Google DeepMind.

The company's lean operational approach has also attracted attention. DeepSeek reportedly achieved its breakthroughs with significantly fewer computing resources than Western competitors, a narrative that captivated global markets earlier in 2025 when its R1 model shocked Silicon Valley. That efficiency story has only strengthened investor appetite.

Tencent and Other Tech Giants Circle the Deal

Tencent, one of China's largest technology conglomerates, is among the investors negotiating for a stake in the round. The social media and gaming giant has been aggressively expanding its AI portfolio, integrating large language models across its WeChat, cloud computing, and enterprise platforms.

However, the final investor roster remains fluid. Sources indicate that multiple parties are still in active discussions, and the composition of the round could change before closing. Key considerations include:

• Strategic alignment: Which investors bring more than just capital — cloud infrastructure, distribution channels, or government relationships
• Founder control: Liang Wenfeng's potential participation in the round suggests he wants to maintain significant ownership and decision-making power
• Regulatory dynamics: Any investment involving the Big Fund carries implicit government endorsement, which could complicate or enhance relationships with other stakeholders
• Geopolitical sensitivity: Western investors and partners may view Big Fund involvement as increasing the strategic significance — and potential restrictions — around DeepSeek's technology

Liang Wenfeng's reported interest in co-investing alongside external backers mirrors a pattern seen in other high-profile AI funding rounds globally. Sam Altman invested in OpenAI's recent rounds, and founders at companies like Anthropic and xAI have similarly maintained or expanded their stakes during major fundraises.

How This Compares to Global AI Funding

DeepSeek's potential $45 billion valuation places it firmly among the world's most valuable AI startups, though it still trails the very top tier. For context:

• OpenAI was valued at approximately $300 billion in its most recent funding round
• Anthropic has been valued at roughly $60 billion following investments from Amazon and Google
• xAI, Elon Musk's AI venture, reached a $50 billion valuation in late 2024
• Mistral AI, Europe's leading AI startup, was valued at approximately $6 billion

A $45 billion valuation would make DeepSeek the most valuable AI startup in China and one of the top 5 globally. It would also dwarf the valuations of other prominent Chinese AI companies like Baidu's Ernie Bot division, Zhipu AI, and Moonshot AI (Kimi).

The scale of this round also reflects the intensifying AI arms race between the U.S. and China. As American companies raise tens of billions from private markets and receive massive cloud computing commitments from hyperscalers, Chinese AI labs need equivalent capital to remain competitive — particularly given restricted access to cutting-edge NVIDIA GPUs due to U.S. export controls.

What This Means for the Global AI Industry

For developers and businesses worldwide, this funding round carries several practical implications. DeepSeek has positioned itself as a credible open-source alternative to proprietary Western models, and a massive capital injection would likely accelerate its model development cadence.

Companies currently building on DeepSeek's open-source models can likely expect continued free access — at least in the near term. The Big Fund's involvement suggests the Chinese government views DeepSeek's open-source strategy as a feature, not a bug, since widespread adoption of Chinese AI models serves broader strategic interests.

For Western AI companies, a well-funded DeepSeek intensifies competitive pressure. The company has already demonstrated an ability to match or approach frontier model performance at a fraction of the cost. With $45 billion in implied resources and government backing, DeepSeek could invest heavily in:

• Next-generation model architectures and training runs
• Expanded research teams and talent acquisition
• Custom hardware and infrastructure development
• International distribution and developer ecosystem building

Looking Ahead: Uncertain Timeline, Clear Direction

The deal has not been finalized, and several variables could still shift the outcome. The Big Fund's internal approval processes, regulatory reviews, and ongoing negotiations with other potential investors all introduce uncertainty around timing.

However, the direction is clear. China's most powerful technology investment vehicle is signaling that AI foundation models deserve the same level of national strategic support as semiconductor fabrication. This parallels — and perhaps responds to — the U.S. CHIPS Act and the Biden and Trump administrations' broader efforts to maintain American AI supremacy.

If the round closes at or near the reported $45 billion valuation, it will represent one of the largest AI funding events of 2025 and a watershed moment for China's AI ecosystem. DeepSeek would transition from a scrappy, efficiency-focused upstart to one of the world's best-capitalized AI companies — with the explicit backing of the Chinese state.

The coming weeks will reveal whether these discussions translate into a signed deal. But regardless of the exact terms, the fact that China's Big Fund is even at the negotiating table with an LLM company marks a fundamental evolution in how Beijing is approaching the global AI race.