←News detail

36Kr · AIUpdated Nov 2 17:39

Xiaohongshu Participates in DeepSeek's Second Round of 105.5 Billion Yuan Financing

AI summary

The Information disclosed that DeepSeek's second financing round amounted to 105.5 billion yuan, with a valuation of 615.5 billion yuan, and nearly 30 investors participating, with Xiaohongshu appearing on the list. In this round, the proportion of state-owned capital investment rose from 2% in the first round to 27%; Tencent invested 10 billion yuan, unchanged from the first round; CATL and Zeng Yuqun together invested 10 billion yuan. Previously, Xiaohongshu had already participated in Kimi's financing of over 1 billion US dollars and appeared among the strategic investors ahead of MiniMax's listing.

Body · AI translation

Everyone wants to invest in Liang Wenfeng.

The Information disclosed DeepSeek's second round of financing: 105.5 billion yuan in scale, a valuation of 615.5 billion yuan, with nearly 30 investors participating. The lineup is astonishingly luxurious.

Amid the long list, an unexpected figure emerges—Xiaohongshu.

Xiaohongshu followed suit and invested in DeepSeek

Speaking of it, there is not much direct connection between Xiaohongshu's founding team and Liang Wenfeng. Mao Wenchao and Qu Fang grew up in the systems of consumer goods, consulting, and business narratives, while Liang Wenfeng is rooted in the world of mathematics, computing power, and quantitative trading. The only coincidence is that both were born in 1985.

Until April this year, Xiaohongshu established Dots, a first-level department centered on AI—upgraded from the Humanistic Intelligence Lab Hi Lab founded in early 2025—with four major segments under it: model R&D, infrastructure, engineering, and product, reporting to the new president Conan. To the outside world, this marks Xiaohongshu's full embrace of AI.

Then in October, Xiaohongshu issued an all-staff letter announcing the integration of the AI first-level department Dots with the company's core infrastructure (Infra) team, further clarifying internal division of labor: Dots takes on the tasks of model implementation and serving products, while Hi Lab continues to do cutting-edge long-term exploration.

This is Xiaohongshu's second major adjustment to its AI organizational structure within half a year. In the letter, Xiaohongshu repeatedly mentioned that "rather than pursuing model capabilities or computing power scale itself, what matters more is the deep integration of models and applications." This may also be one of the drivers behind Xiaohongshu's investment in DeepSeek.

Externally, Xiaohongshu has acted even more decisively. Leafing through the shareholder registers of Moonshot AI's Kimi and MiniMax, Xiaohongshu's name appears—in 2024 it participated in Kimi's over 1 billion USD financing; before MiniMax's listing, Xiaohongshu likewise appeared among the strategic investors.

Add this round's DeepSeek, and Xiaohongshu seems to have secured all the "tickets" to China's first tier of large models.

Hundred-billion financing

This is a list Liang Wenfeng carefully considered.

According to The Information, DeepSeek's second round of financing was originally as high as 105.5 billion yuan, double the size of the first round; the share of state-owned capital jumped from just 2% in the first round to 27%. Among them, the National Artificial Intelligence Industry Investment Fund continued to add 1 billion yuan, while China Mobile, China Telecom, CICC, CITIC, and China Life all joined in, with investment sizes ranging from 2 billion to 8 billion yuan.

Industrial capital is even queuing up to grab shares.

The Information's report shows that in this round, Tencent and CATL remain the two largest investors. Tencent invested 10 billion yuan this round, flat with the first round, for a cumulative 20 billion yuan across both rounds; CATL and its chairman Zeng Yuqun together invested 10 billion yuan this round, double the 5 billion yuan of the first round.

The new investor lineup is nothing short of luxurious: on the chip side, SMIC and ChangXin Memory Technologies; on the manufacturing side, Luxshare Precision and Midea; among automakers, Geely; among internet giants, Meituan, vivo, and Ctrip; plus Xiaohongshu—rumors say each of the major companies invested 1 billion to 3 billion yuan.

On the market-oriented fund side, Monolith Management invested 3 billion yuan and 7.5 billion yuan in the two rounds respectively; IDG Capital invested a cumulative 6.5 billion yuan across both rounds; Boyu Capital invested 3.5 billion yuan in the second round.

In DeepSeek's first round of about 50 billion yuan in June this year, founder Liang Wenfeng himself contributed 20 billion yuan, making him the largest single investor at the time. In the second round, he again contributed 2 billion yuan.

To maintain the independence of the company's R&D and the purity of decision-making, DeepSeek set a special threshold in its first round. Apart from the National Artificial Intelligence Industry Investment Fund, other external investors' funds do not go directly into the DeepSeek entity but instead enter uniformly through a limited partnership managed by Liang Wenfeng—giving up voting rights and committing to lock up shares for five years.

That all parties keep coming forward is, to some extent, a kind of era anxiety: the next-generation super entry point will most likely be born in AI. Investing in DeepSeek not only means obtaining a ticket to the future but also opening a door to reshaping one's own business.

This article comes from the WeChat public account "Investment World" (ID: pedaily2012), author: Zhou Jiali, published by 36Kr with authorization.

Source

36Kr · AI