DeepSeek’s first-round financing raised $7.35 billion, outpacing Alibaba! It’s only the money with the “minimum conditions.”

China’s AI startup DeepSeek has launched its first round of external financing since it was founded. It plans to raise up to 7.35 billion USD (500 billion RMB), with a post-investment valuation expected to exceed 51.5 billion USD (350 billion RMB). This would make it the largest single financing deal in China’s AI history. Alibaba’s negotiations collapsed after it demanded ecosystem linkage. Tencent’s proposal to subscribe for 20% was also declined. Founder Liang Wenfeng plans to personally invest up to 200 billion RMB, subscribing for roughly 40% of this round.
(Background: Huang Renxun: Blackwell and Rubin should not be allowed to let China obtain them; even though NVIDIA China’s market share is zero, it still refuses to back down.)
(Additional context: Anthropic spent 200 billion USD over five years on Google Cloud; the two AI startups have eaten up half of the orders of four major cloud giants.)

Table of Contents

Toggle

• Key Highlights
• Alibaba Rejected, Tencent Cut, National Team Expected to Lead
• V4.1 to Launch in June, but Core Researchers Have Already Been Poached
• The Pure Research Mode Has Reached Its Limit
• Frequently Asked Questions

Key Highlights

• DeepSeek’s first external funding round targets up to 7.35 billion USD (500 billion RMB); founder Liang Wenfeng personally invests 200 billion RMB, accounting for 40%
• Alibaba’s ecosystem-linkage demands caused negotiations to fall apart; China’s National Integrated Circuit Industry Investment Fund (Big Fund) is expected to lead with the most relaxed terms
• V4.1 is expected to be released in June with native support for MCP, but core researchers have already moved to Xiaomi and ByteDance

DeepSeek, which has not touched external financing for three years, has officially opened the door. According to The Information, this Chinese AI startup incubated by the quant fund Fantom Quant has launched its first external financing round, aiming to raise 500 billion RMB (about 7.35 billion USD). Its post-investment valuation is expected to exceed 3,500 billion RMB (about 515 billion USD). If completed, it would be the largest single financing in China’s AI industry history.

But even with the door open, what founder Liang Wenfeng wants is “money with the fewest conditions.”

Alibaba Rejected, Tencent Cut, National Team Expected to Lead

The biggest investor in this round is not an external institution, but Liang Wenfeng himself—the founder. He plans to personally invest up to 200 billion RMB, subscribing for about 40% equity. China’s National Integrated Circuit Industry Investment Fund (Big Fund Phase 3’s AI special project) is expected to become the second-largest investor, with relatively more relaxed terms.

Alibaba and Tencent both actively negotiated, but both hit a wall. Alibaba hoped to strengthen its own AI ecosystem flywheel by deeply integrating business units such as Tongyi Qianwen into product lines like Taobao and Amap. DeepSeek insists on technical independence and does not accept ecosystem linkage, and the two sides ultimately failed to reach an agreement.

Tencent proposed subscribing for up to 20% of shares, but it was declined because the dilution was too large.

Liang Wenfeng has long refused external equity financing. Previously, DeepSeek was fully supported by Fantom’s own funds, with Liang’s direct and indirect holdings totaling about 84%. Although this is the first time it has opened the door, the bottom line is an offer with “the fewest add-on conditions.” The Big Fund’s terms are the loosest and most aligned with DeepSeek’s path.

V4.1 to Launch in June, but Core Researchers Have Already Been Poached

The commercial timeline pushed by the financing is accelerating. DeepSeek plans to release a V4.1 update in June. Key efforts will focus on strengthening enterprise-side tools, providing native support for MCP (Model Context Protocol), and adding image and audio processing capabilities (generation on the producer side remains limited to text). This is DeepSeek’s first clear move to target the enterprise market with tool-level products.

But talent loss is more urgent than the schedule. A core contributor to V3, Luo Fuli, left for Xiaomi in November last year and took over the MiMo large model team. By April this year, MiMo-V2.5 released under his leadership has already surpassed DeepSeek-V4-Flash in multiple evaluations. Another core researcher, Guo Dayā, joined ByteDance’s Seed team in April this year as the head of the Agent direction (L8 level). There are rumors that his salary is “close to 100 million RMB” (ByteDance denies specific figures). Reports indicate ByteDance has poached nearly 70 people from DeepSeek over the past year.

The salaries offered by competitors are more than twice what DeepSeek pays. One major use of the new funds is to raise employee compensation and grant stock options. This creates market-based pricing for equity through external financing, giving liquidity to paper wealth.

The Pure Research Mode Has Reached Its Limit

In February this year, DeepSeek also publicly said, “Fully funded by Fantom, with no plan for immediate fundraising.” Three months later, skyrocketing compute costs and a brutal talent war forced it to change its stance.

To compare with U.S. model vendors, the combined cloud compute contracts of Anthropic and OpenAI are close to 2 trillion USD. DeepSeek’s reputation for training cutting-edge models at extremely low cost is certainly impressive, but maintaining technical leadership requires not only efficiency, but also absolute scale of compute power and compensation that can retain talent.

By embracing funding from the national team, launching enterprise-grade tools, and pricing employee equity, DeepSeek is facing the need to self-generate capital and enters an even larger global-level competition.

Frequently Asked Questions

How big is this round of funding for DeepSeek?

Up to 500 billion RMB (about 7.35 billion USD), with a post-investment valuation exceeding 3,500 billion RMB (about 515 billion USD). This is the largest single financing deal in China’s AI history. Founder Liang Wenfeng personally contributes about 40%.

Why couldn’t Alibaba invest in DeepSeek?

Alibaba wanted to embed DeepSeek’s technology into its own product ecosystem (including Taobao and Amap), but DeepSeek insisted on technical independence and minimizing terms that would bind it. Negotiations between the two sides fell apart. Because its terms are the most relaxed, the Big Fund is expected to lead the investment.