In early June, Reuters reported, citing sources familiar with the matter, that DeepSeek is planning to raise approximately $7 billion (around 50 billion yuan) in its first external funding round, with a post-money valuation potentially reaching $59 billion (approximately 400 billion yuan).
This is not an ordinary fundraising event.
It marks the first time DeepSeek has accepted external investment in nearly three years since its founding.
Just over a year ago, Liang Wenfeng was publicly stating the company's principles: "no fundraising, no IPO, no commercialization."
These three ironclad rules have all been broken within a span of three months.
Breaking the Three Ironclad Rules
In July 2023, Hangzhou DeepSeeker Artificial Intelligence Basic Technology Research Co., Ltd. was registered and established.
Liang Wenfeng established three ironclad rules for DeepSeek: no fundraising, no IPO, no commercialization.
During the 2023-2024 period when capital was flooding into the AI sector, this was an almost "rebellious" choice.
Liang Wenfeng's reasoning was straightforward: DeepSeek's goal is to build a world-class general-purpose large language model, not to make money or go public. He believed capital would demand short-term returns and commercialization would compromise the technical roadmap, both of which would interfere with the company's core objectives.
This statement circulated within the AI community for three full years.
However, in April 2026, things began to change.
First, industry rumors emerged that DeepSeek was in talks with Alibaba and TENCENT for its first external equity financing, with valuation discussions reaching $20 billion. Subsequently, at the end of May, the V4 model sparked a market frenzy, driving the valuation to nearly $60 billion. Then, in early June, Reuters broke the significant news: a first funding round of $7 billion at a $59 billion valuation, led by TENCENT and Contemporary Amperex Technology Co.,Ltd. (CATL).
The three ironclad rules were all broken within three months.
Source of the Funds: A $20 Billion Founder Contribution
The structure of this funding round is particularly interesting.
Several noteworthy details stand out.
First, Liang Wenfeng personally contributed 20 billion yuan.
This is not a simple case of "I'll invest too"; it is a lead investment.
A founder personally investing 20 billion yuan to lead a funding round for his own company is extremely rare in the AI sector. One interpretation is that it is to price the team's stock options. DeepSeek needs a market-validated valuation to establish a reasonable price for employee stock options.
Second, the total number of investors is strictly limited to under ten.
This is not a case of "whoever has money can join"; it is a selective process. TENCENT, Contemporary Amperex Technology Co.,Ltd., NetEase, JD.com—all are industrial capital, with no pure-play financial venture capital firms.
Third, NetEase and JD.com are still in the negotiation stage.
This means the final list of investors may still change.
Reasons for the Strategic Shift: A Calculated Response to Pressure
Liang Wenfeng did not suddenly "sell out." He made a carefully calculated choice in the face of three major pressures.
Pressure one: Talent drain (immediate concern). Multiple key technical personnel have been poached with high salaries. A direct purpose of the fundraising is to price employee stock options and use equity to retain talent.
Pressure two: Computing power and funding pressure (long-term concern). The AI race has evolved from algorithmic ingenuity to a "naked war of computing power consumption." The upcoming V4.1 release requires massive computing power investment.
Pressure three: Industry competition and the valuation window. The soaring valuations of companies like Zhipu AI and MiniMax post-IPO have set benchmarks. If the window is missed, private market valuations could be "instantly shattered."
Strategic Implications of Key Investors
The entry of TENCENT and Contemporary Amperex Technology Co.,Ltd. carries symbolic significance far beyond the monetary investment.
TENCENT's involvement represents a strategic move, both defensive and offensive. Its self-developed "Hunyuan" large model currently lags behind ByteDance's "Doubao" and DeepSeek domestically. Establishing a strategic partnership with DeepSeek helps address this gap. This is a strategic investment, not merely a financial one.
Contemporary Amperex Technology Co.,Ltd.'s involvement signals that the AI competition is extending from algorithms to energy infrastructure. CATL is actively entering the AI data center field, exploring power equipment and energy storage solutions. The next battlefield for AI is not just algorithms, but energy. CATL's participation means DeepSeek has secured a strategic partner at the energy infrastructure level. This is an ecosystem investment, not just a financial one.
Impact on Developers: Positive or Negative?
Is DeepSeek's $7 billion fundraising good or bad news for developers?
On the positive side, API prices may remain low. DeepSeek's current pricing is about one-tenth that of competitors. The $7 billion funding means DeepSeek has the capacity to continue "subsidizing" developers and maintain its low-price strategy. The open-source strategy may also continue, as Liang Wenfeng's significant personal investment aims to protect control and avoid pressure from financial VCs to pursue an IPO or commercialization aggressively. Furthermore, technical iteration may accelerate with more funds for computing power, talent retention, and R&D.
On the negative side, the promise of "no commercialization" has been broken. Investors who contributed $7 billion are not philanthropists; they need an exit path, which necessitates eventual commercialization. API prices may not stay this low forever, as investors will seek returns. There is also a possibility that the open-source strategy could change if future investors demand some models be closed-source to protect commercial interests.
Future Outlook: IPO, Open Source, and API Pricing
Following the $7 billion funding, three key questions warrant attention.
The first is whether an IPO will happen. Liang Wenfeng has stated there is "no rush to go public," but this was a statement from early June. With $7 billion invested and a $59 billion valuation, investors will require an exit path. An IPO seems a matter of timing, but "no rush" might imply it won't happen before 2027.
The second is whether the open-source strategy will change. The current commitment to "continue open-sourcing" is based on Liang Wenfeng's control. However, if the company goes public in the future, investors might demand some models become closed-source.
The third is whether API prices will increase. In the short term, likely not, as the funding provides a cushion. Long-term, however, prices may gradually rise to achieve commercialization.
Concluding Analysis
DeepSeek's first funding round of $7 billion at a $59 billion valuation is not an ordinary financing event.
It represents Liang Wenfeng's second major choice, shifting from "technological idealism" to "calculated realism."
The first choice was in July 2023, establishing the "three ironclad rules."
The second choice was between April and June 2026, breaking those rules and accepting external funding.
For developers, this means short-term benefits: maintained low API prices, continuation of the open-source strategy, and accelerated technical iteration.
It also brings long-term uncertainties: commercialization pressure, potential API price hikes, and possible changes to the open-source policy.
Liang Wenfeng says there is "no rush to go public."
But the market will say: "Once the money is in, you are accountable to it."
Comments