Introduction
Reliable news reports published by economic agency Bloomberg on Wednesday stated that Chinese artificial intelligence startup DeepSeek has revived its second external funding round following a hiatus of several weeks. Through this intensive round, the ambitious technology company seeks to secure massive funding approaching the $8 billion mark, based on a massive and staggering total company valuation of approximately 500 billion Chinese yuan, roughly equivalent to $74 billion according to current exchange rates. According to accompanying economic reports, prominent Hong Kong-based asset management firm Monolith Management is currently in advanced and serious talks to participate heavily in this massive investment round. The Hangzhou-based company had temporarily suspended this process last month following frustration and confusion resulting from leaked remarks made by founder and CEO Liang Wenfeng to investors in closed-door sessions. The resumption of the current round represents a rapid and massive increase in financial valuation compared to the company’s first funding round held last June, which successfully raised about 50 billion yuan with generous backing from the state-owned China National Artificial Intelligence Industry Investment Fund, alongside a group of major tech companies such as Tencent, CATL, NetEase, and JD.com.
- Data center ambitions and chip challenges
- Price increases and the drive toward profitability
- Intensive preparations for an initial public offering
- Frequently asked questions
Data center ambitions and chip challenges
The financial proceeds gained from this massive round are expected to be directed toward funding a major and vital expansion plan in the company’s AI infrastructure, specifically including the construction of a large-scale, high-tech data center in the vast Inner Mongolia region. This vital strategic project aims to secure massive and sustainable computing capacity reaching 1 gigawatt of power. This intensive, capital-intensive expansion highlights a broader strategic shift across China’s AI sector, where competition is rapidly shifting from mere superiority in model software benchmarks to a fierce struggle to secure the raw computing power and physical resources needed to run massive servers.
On a parallel front, DeepSeek faces significant restrictions and challenges in purchasing advanced electronic chips due to strict export controls imposed by the United States government, which severely limit Chinese companies’ access to the latest and most efficient processors produced by global leader Nvidia. In this complex context, multiple intersecting reports indicate that the company is currently working intensively and secretly on developing its own inference chip to meet its growing needs, as part of China’s relentless and ongoing efforts to find domestically made technological alternatives capable of competing with Western technology and reducing external dependence.
Price increases and the drive toward profitability
The revival of ambitious funding plans coincides with a significant and decisive announcement issued by DeepSeek stating that it will comprehensively and effectively raise API service prices across its digital platform. The company clarified in its statement addressed to clients and developers that this increase will be “significant and tangible,” although it has not yet disclosed exact figures and percentages. Under current model pricing known as V4 Flash, the company charges nominal and discounted fees of $0.14 per million input tokens and $0.28 per million output tokens, a competitive and aggressive pricing scheme well below the rates charged by prominent global market competitors such as Anthropic and Moonshot.
This strategic move to raise prices comes on the heels of a massive surge and unprecedented demand following the launch of its newest model last month, which pushed the platform’s daily output to trillions of tokens processed in real time. Technical developer platform OpenCode reported that this innovative model has already processed 8 trillion tokens in a single day, specifically on August 1st, reflecting immense pressure on the company’s servers and the necessity of generating revenue commensurate with high consumption volume to ensure continuity.
Intensive preparations for an initial public offering
In another related development, DeepSeek is preparing at an accelerated pace for the potential execution of an initial public offering of its shares on the prestigious STAR Market in Shanghai, China. Previous economic reports published by Reuters had clearly indicated that the current fundraising process was partially designed to position the company in a strong, competitive financial standing that makes it attractive to investors and financial markets ahead of its planned mainland China listing. Bloomberg noted in its in-depth analysis of the economic situation that the Chinese tech giant faces “three major challenges simultaneously in the upcoming period that cannot be overlooked: managing extremely high investor expectations, expanding rapidly into new markets to secure a larger market share, and building complex computing infrastructure that requires massive financial investments that are indispensable in the global AI race.”
Frequently asked questions
Question: What is DeepSeek’s target valuation in the current funding round?
Answer: The company is striving to secure funding approaching $8 billion, based on a total valuation estimated at around 500 billion Chinese yuan, roughly equivalent to $74 billion, reflecting massive and rapid growth for the tech startup.
Question: Where will these massive funds be primarily invested after they are raised?
Answer: These multi-billion dollar investments will be directed primarily toward expanding the company’s technological infrastructure, most notably building a giant data center in Inner Mongolia to secure stable computing capacity reaching 1 gigawatt to support advanced AI research and applications.
Question: How is the company dealing with strict US export controls on smart chips?
Answer: To overcome this technical obstacle that directly threatens its operations, the company has secretly begun developing its own inference chips and processors locally as a vital alternative to the advanced processors that Washington bans from export to Chinese companies operating in this strategic field.
Question: Why did the company decide to raise its API service prices at this time?
Answer: This decisive decision comes as a result of a very significant increase in demand and consumption volume following the launch of its latest model, which processed 8 trillion tokens in a single day. This step is considered a serious attempt to move toward sustainable profitability and cover the high operational costs of servers.
