The DeepSeek Financing Story

@Jing Liu

DeepSeek financing news is everywhere by now.

What we already know, elsewhere won't rehash. Below are some untold stories and details we've learned.

The Legendary Four-Hour Meeting

First, that investor meeting — the one passed around by word of mouth as "the four-hour meeting."

In mid-May, DeepSeek organized this virtual session over Tencent Meeting. By then, the investor lineup was more or less set, with two slots per firm.

Wenfeng Liang spoke first, followed by a Q&A session where investors asked questions and he responded.

Legend has it that one CVC investor spent a full ten minutes on his questions alone: a long self-introduction, followed by three lengthy questions. The atmosphere grew awkward at one point — not hard to imagine. People whispered among themselves, asking who this person was.

But Liang didn't seem bothered, earnestly answering every question. Though after the second answer, he momentarily forgot what the third question was and had to ask the investor to repeat it.

Before this, even most of those who had already committed capital hadn't met Liang in person. This was their first real exchange with him. After the meeting ended, many were left with a lingering sense of something more — a feeling that stayed with them.

Liang's memorable lines and spirit began circulating in the market.

Restraint and AGI

Wenfeng Liang isn't an exceptionally articulate person, nor does he speak quickly. But several quotes from that meeting are often recalled (paraphrased, the original wording may vary slightly):

"We have no fame and no pull. We are a group of very ordinary people."

"The narrative I love is: ordinary people doing extraordinary things. For DeepSeek, our core priority is ensuring team stability — this matters more than money or resources. This is our biggest risk, and our biggest challenge."

"Assume AGI will account for 20% of GDP. Those who want 5% will lose to those who want 1%. Those who want 1% will lose to those who want 0.1%."

"We will only do things related to advancing intelligence. We won't do anything else."

"AGI is a big enough thing. Everything else is just process."

Through all four hours, Liang kept returning to the same themes: the sole pursuit of AGI; less is more; restraint.

From 5 Billion to 1.5 Billion

Formal fundraising kicked off around April this year. There was reportedly a "fund whitelist": institutions with real capital capacity and strong brands.

Initially, DeepSeek required a minimum commitment of 5 billion RMB, no syndication of shares, and a pure RMB structure.

Against this backdrop, one investor figure who was initially rumored as a certain participant but ultimately didn't make the cut — widely speculated to be because he had once tried to sell allocation to Middle Eastern LPs.

But in China, there are very few market-oriented funds capable of writing a 3 billion RMB single check. So some VCs repeatedly pushed: could they participate through multiple fund vehicles?

In the end, the minimum fund commitment was lowered to 1.5 billion, and fundraising structures became more flexible.

There was also talk that participating institutions preferably shouldn't have invested in other large language model companies. This was likely impossible — the contradiction being obvious: first-tier funds that haven't touched LLMs are rare.

At an investor meeting, Wenfeng Liang also expressed that DeepSeek is willing to help other companies — even large model companies. The condition, of course, remains that they don't poach his people.

Monolith

Monolith appeared on investor lists relatively early. What changed was the number: initially 1.5 billion, ultimately 3 billion.

One explanation for the shift ties to allocation adjustments with Guozhi Investment, freeing up some capacity — the latter's 980 million RMB figure suggesting an intentional rounding down below the 1 billion threshold.

Either way, making bold bets is distinctly Xi Cao's style. He proved this once before with Moonshot AI.

Monolith is likely the youngest VC among participants in this round.

Loyal Valley Capital

Loyal Valley Capital may be the most surprising name on the final list.

Many have interpreted Loyal Valley Capital's move as a case of FOMO — having sat out China's large-model wave entirely, they're desperate to get in. But from what we understand, they actually did invest in Moonshot AI, at a valuation somewhere between $6 billion and $10 billion. Relative to their overall fund size, however, the check was fairly modest.

One market interpretation is that Loyal Valley Capital's founder comes from a secondary-market background, which supposedly creates a natural connection with High-Flyer Quant. But in reality, quantitative investing and fundamental active management operate in almost entirely separate circles — and often represent fundamentally different value systems.

As we understand it, this investment came together not only through Lijun Lin and his partner Chunyan Ye, but also through the efforts of a newer investment hire.

Loyal Valley Capital was also among the earliest VCs to start talking with DeepSeek.

Could DeepSeek Be Hillhouse and HSG's Biggest Miss?

Now for the most surprising part of this deal.

Initial rumors had Hillhouse participating, and committing 5 billion. Lei Zhang had reportedly told friends multiple times that since 2025, he and Liang had been in regular contact, and that he'd visit Liang's home when in Beijing.

But this gradually shifted to: Hillhouse wouldn't participate, but Lei Zhang might invest personally. Reasons cited ranged from Hillhouse's fundraising timeline to regulatory concerns.

In the end, neither Zhang nor Hillhouse appeared on the list.

HSG, on the other hand, was widely seen as unlikely to participate almost from the start. The most common explanation making the rounds was concern about its overseas LPs. We raised this with several people close to HSG, and they broadly conveyed the same sentiment. Of course, other theories have been circulating too.

The upshot: HSG and Hillhouse — two institutions that seemed impossible to exclude — both ended up absent. A top lawyer told us that with a RMB-structured company, hiding participation is theoretically impossible.

Among established VCs, only IDG invested.

So someone turned the question back on us: Could DeepSeek end up being the biggest miss for Sequoia and Hillhouse?

Hard to say. But as we understand it, DeepSeek may not stop at this single round.

100 Institutions and Individuals

The investor list shows only 10 participants. But elsewhere did a quick look through fund structures, and found considerably more actual parties involved.

For example, the external CVC of the CATL ecosystem — Puquan Capital's backers include: Xiamen and Ordos local government capital, CATL-related entities, and the National Green Development Fund.

IDG Capital funds involved in this transaction have a distinctly insurance-funded character, accounting for a very high proportion of the 3 billion yuan. In addition, entities related to Nongfu Spring and Septwolves are also participating.

Monolith's current LPs are primarily Zhejiang and Shanghai local government capital, plus listed companies like By-Health and Yanghe Brewery.

Furthermore, iHealth (Andon Health) is also among participants. iHealth is in fact an LP behind numerous GPs.

A rough estimate, with shallow penetration and simple deduplication: approximately 100 institutions or individuals participated in this DeepSeek financing round.

The Top Demand: No Poaching

DeepSeek's investment terms won't be rehashed here. Just one additional point: whether for Big Tech or VC funds, Wenfeng Liang's most important requirement is this — don't poach DeepSeek's people, or encourage them to leave and start their own companies.

"China's Largest Listed Company"

If we view this as a pure investment, what kind of bet is DeepSeek?

One participating investor once worried aloud: could this be an overly consensus investment? In investing, consensus outcomes tend to be mediocre.

But for DeepSeek, it's hard to say that. Or rather, we can't understand it through a simple investment lens. Whether for its significance to China's AGI industry, or its symbolic meaning.

As one investor put it: "They're doing all this with enormous goodwill. He seems to be telling us: this is how the world should be."

So some hesitated, while others believed — "there's no reason not to invest." If, as some expect, an IPO materializes, could DeepSeek become the largest listed company on China's A-shares?

Quiet Ambition

In following DeepSeek's financing, one striking impression: for most people, DeepSeek was an unexpected answer. Even the most competitive investors never thought it would open up to fundraising.

So as mentioned, most well-known investor figures only connected with the company recently. Objectively, Liang had long avoided meeting investors.

When financing news broke, many institutions didn't even try to pitch for allocation. Yet in the end, some got tickets through claimed synergies, some through conviction, some through brand alignment, some simply through effort.

When elsewhere tried asking some investors about this deal, most didn't want to say much. The sensitivity of the transaction was only part of it. More than that, there seemed to be a kind of quiet pride — a sense of being associated with something notable — that made them want to keep things low-key.

One investor said he treasured every document throughout the process, hoping to frame them someday. Another repeated twice the sixteen characters from the DeepSeek V4 preview release: "Not tempted by praise, not frightened by slander. Follow the path, correct oneself."

Cover image: Joseph Mallord William Turner, Whalers, 1845, Metropolitan Museum of Art