Chinese artificial intelligence start-up DeepSeek is finalising its first external funding round of more than 50 billion yuan, equivalent to about $7.4 billion or £5.5 billion, at a valuation just under $60 billion. The round represents a more than six-fold increase from a $10 billion valuation reported only two months earlier, according to people familiar with the matter. The rapid revaluation underscores how quickly investor appetite for leading AI developers has grown, particularly in China, where DeepSeek has become a symbol of cost-efficient model development and domestic technological ambition.
The reported investor list includes some of China’s largest technology and industrial companies. Tencent Holdings is expected to invest 10 billion yuan, while NetEase and JD.com are each expected to contribute about 3 billion yuan. CATL, the world’s largest electric vehicle battery manufacturer, is expected to contribute about 5 billion yuan. Venture capital firms including IDG Capital, Monolith, Loyal Valley Capital and Shixiang Tech are also reportedly expected to participate. DeepSeek founder and chief executive Liang Wenfeng is expected to contribute about 20 billion yuan of his own capital, a move that would reinforce his control and commitment to the company.
State-backed funds and final terms
State-backed funds are also expected to take part in the round, although final terms have reportedly not been settled. Those unresolved details include which investment vehicles will provide the capital and how the funding will be structured. The involvement of state-linked investors would align with Beijing’s broader push to support domestic AI champions and reduce reliance on foreign technology. It would also give DeepSeek access to capital and policy support as it scales its models and expands its ecosystem.
DeepSeek is backed by High-Flyer Quant, the highly profitable quantitative fund led by Liang. The start-up previously resisted bringing in external capital, preferring to grow with backing from its founder’s quant fund. More recently, however, it decided to open its first external funding round. Local media reports indicate that one motivation is to secure a well-defined market value that can help the company retain top talent. In the global AI race, equity compensation and clear valuations have become important tools for keeping researchers and engineers from being poached by rivals.
Why the valuation jump matters
The jump from $10 billion to just under $60 billion in two months would make DeepSeek one of the most valuable private AI companies in the world. It also highlights a broader trend: investors are willing to pay premium prices for AI developers that can demonstrate technical breakthroughs, commercial potential and a credible path to scaling. DeepSeek first disrupted the AI world last year with the release of powerful, cost-efficient models. Those models challenged the assumption that only companies with massive compute budgets could train and deploy state-of-the-art AI. By optimising training and inference, DeepSeek showed that smaller, more focused teams could compete with larger players on key benchmarks.
The company has continued in the same vein with recent releases. When it released its V4 model in April, DeepSeek said prices would “drop significantly” in the second half of this year. That projection was tied to the expected large-scale shipment of Huawei’s Ascend 950PR supernodes. If those chips ship in volume, they could lower the cost of running advanced AI workloads in China and reduce dependence on imported hardware. That dynamic is central to DeepSeek’s strategy: cheaper models and cheaper compute can expand adoption across industries, from consumer applications to enterprise software, research and public services.
Domestic chip ecosystem
Chinese chipmakers including Huawei, Cambricon and MetaX have said they are optimising their AI processing chips for DeepSeek’s V4. The work is part of China’s drive to boost its locally developed AI supply chain. For DeepSeek, compatibility with domestic chips could provide a hedge against export controls and supply chain disruptions. For chipmakers, DeepSeek’s models offer a high-profile workload that can demonstrate the capabilities of their hardware. The collaboration is mutually reinforcing: better chip support makes DeepSeek’s models more efficient, while DeepSeek’s popularity gives domestic chip designers a flagship customer and a reason to accelerate their roadmaps.
The funding round also comes amid intense global competition in generative AI. Companies around the world are racing to build larger models, deploy agents and integrate AI into products. The cost of compute, energy and talent has risen sharply. Access to capital is therefore a strategic advantage. DeepSeek’s reported round, if completed, would give it a war chest to invest in research, infrastructure, hiring and partnerships. It could also help the company negotiate better terms with suppliers and expand its commercial efforts.
Investor mix signals confidence
The mix of investors is notable. Tencent, NetEase and JD.com bring consumer internet expertise, distribution channels and enterprise relationships. CATL brings industrial scale and deep ties to China’s advanced manufacturing sector. Venture capital firms bring experience in scaling technology companies. The participation of Liang himself, expected to contribute about 20 billion yuan, would be one of the largest founder commitments in a Chinese AI funding round. It signals that the founder sees long-term value in the company and is willing to concentrate more of his personal capital in it.
At the same time, the reported round is not yet final. The size and valuation are based on people familiar with the matter, and final terms could change. State-backed funds may join, but the vehicles and amounts are still being discussed. Such uncertainty is common in large private financings, especially when strategic investors, financial investors and government-linked entities are involved. Each group may have different objectives: financial returns, strategic access, supply chain coordination or national technology goals.
Talent and market value
One of the most important reasons for the round, according to local media reports, is talent retention. DeepSeek’s success has made its researchers highly sought after. In AI, top engineers and scientists can command enormous compensation packages. A clearly established valuation allows a company to offer equity that employees can value more confidently. It also helps in recruiting, because candidates can assess the potential upside of joining a company with a recognised market value. By raising external capital, DeepSeek can create a market-tested valuation without necessarily giving up control, especially with the founder contributing a large share.
The company’s cost-efficient approach has also shaped its reputation. DeepSeek’s models are designed to deliver strong performance at lower cost, which matters for widespread adoption. In many industries, the barrier to AI adoption is not just model quality but also the expense of running inference at scale. If DeepSeek can lower prices significantly in the second half of the year, as it has suggested, it could accelerate use cases in customer service, coding, education, healthcare and public administration. Lower prices could also pressure competitors to reduce their own prices, benefiting users but squeezing margins.
Global context and competitive pressure
The AI investment landscape has become increasingly polarised. A small number of companies attract huge valuations, while others struggle to raise follow-on funding. DeepSeek’s reported valuation places it among the elite. Its rise also reflects China’s determination to build a self-reliant AI ecosystem. That ecosystem includes not only model developers but also chip designers, cloud providers, data centres and application companies. Government support, state-backed funds and large corporate investors are all part of the effort to create alternatives to foreign hardware and software stacks.
For Tencent, NetEase, JD.com and CATL, investing in DeepSeek offers exposure to a strategic technology. Tencent has its own AI efforts, but backing DeepSeek can provide access to models and talent. NetEase and JD.com can use AI to improve content, e-commerce, logistics and customer engagement. CATL may see AI as increasingly important for battery research, manufacturing optimisation and energy management. Venture firms are seeking financial returns from a company that has already proven it can generate global attention.
The final composition of the round could have implications for DeepSeek’s governance and strategy. If state-backed funds take significant stakes, the company may face expectations around domestic priorities, data security and industrial policy. If private investors dominate, commercial expansion may be the primary focus. The founder’s large personal contribution could give him continued influence over decisions. The balance among these groups will matter as DeepSeek invests in compute, hires talent and builds partnerships.
DeepSeek’s V4 model is central to its near-term roadmap. The company has tied price reductions to the availability of Huawei’s Ascend 950PR supernodes. If those systems ship at scale, they could provide a domestic alternative for AI training and inference. The performance and reliability of those supernodes will be closely watched, because they could determine whether Chinese AI companies can sustain rapid progress under export restrictions. The collaboration with Huawei, Cambricon and MetaX suggests that DeepSeek is preparing its software stack for a multi-chip future.
Ultimately, the funding round would give DeepSeek more resources to compete at a time when AI development costs are rising. It would also provide a clear valuation for employees, investors and partners. The reported investor list reads like a cross-section of China’s technology and industrial elite: internet platforms, e-commerce, gaming, batteries and venture capital. Their participation would signal confidence in DeepSeek’s model efficiency, commercial potential and role in the domestic AI supply chain. Yet the final terms remain unsettled, and the company has not confirmed all details. Until the round closes, the exact valuation and investor allocations may continue to shift.
Chinese chipmakers including Huawei, Cambricon and MetaX are optimising their AI processing chips for DeepSeek’s V4, part of a broader national effort to strengthen a locally developed AI supply chain.
Source: Silicon UK News