DeepSeek discusses its first round of external funding, valuation at $20 billion: China’s AI valuation hits a new high

According to a Bloomberg report dated April 22 (citing exclusive coverage by The Information), Chinese AI startup DeepSeek is in talks for its first round of external fundraising, with a valuation of $20 billion. This marks DeepSeek’s first time raising funds from outside since it was founded in 2023; in the past, it was fully funded internally by the quant hedge fund High-Flyer Capital Management. The $20 billion valuation is also a milestone for Chinese AI startups, as it represents the first time DeepSeek has entered the latter half of the “$1 billion-plus valuation” bracket.

Fundraising size and intended use of funds

DeepSeek is seeking at least $300 million in its first round of funding. The $20 billion valuation doubles the “valuation of more than $10 billion” disclosed earlier on April 17 by The Information. The identities of the investors have not been made public yet, but market speculation suggests it includes funds related to China’s sovereign wealth and international technology investors.

There are three main intended uses for the funds: first, to establish a formal equity incentive program to retain core talent—recently, DeepSeek engineers have continued to be poached by Chinese tech giants such as Tencent and Alibaba with high salaries; second, the training costs for the next generation of large models are surging, and the V4 model and the subsequent V5 both require large-scale GPU resources; third, to strengthen its commercial product lines (API stability, enterprise user support, and ecosystem partners).

A signal that the valuation doubled within a week

When The Information first disclosed DeepSeek’s fundraising talks on April 17, the valuation was “more than $10 billion.” Just five days later, Bloomberg cited The Information to update the figures, and the valuation jumped to $20 billion. It’s not hard to guess why the valuation doubled within a week: DeepSeek’s V4 (a trillion-parameter model) is set to be released at the end of April, the narrative of China’s AI catching up to mainstream U.S. models is gaining momentum, and, in addition, OpenAI/Anthropic’s valuation ranges—$380 billion for OpenAI and $380 billion for Anthropic—provide a point of comparison, making investors willing to pay a higher premium for China’s leading AI company.

Compared with valuations of other Chinese AI startups in the same period: Moonshot AI (the parent company of Kimi) was in the $18 billion range in March; MiniMax and Zhipu AI were roughly in the $5–$10 billion range. DeepSeek’s $20 billion valuation places it directly into the first tier of Chinese AI startups, second only to the AI unit valuations of established giants such as Baidu, Alibaba, and ByteDance.

A unique path from quant funds to an AI unicorn

DeepSeek’s rise follows a path different from that of most AI startups. The company was incubated internally by the Chinese large quant hedge fund High-Flyer Capital Management, and the founder Liang Wenfeng is also a founder of High-Flyer. This meant that in its early stage, DeepSeek relied entirely on High-Flyer’s internal cash flow and GPU resources, without needing to raise funds from outside.

This “quant fund built-in GPU cluster → shift to AI research → use open-source models to build its reputation” path has been imitated by multiple Chinese AI teams between 2025 and 2026, forming a distinctive “quant + AI” integrated structure. But as the training cost for V4 rises (estimated to exceed $100 million per training run) and competition for talent intensifies, DeepSeek has finally moved toward raising funds externally, also marking its organizational transformation from an “internal project” to an “independent company.”

China’s push to catch up to the U.S. in AI

For the international AI industry, DeepSeek’s rise shows that Chinese AI can still enter the global developer ecosystem with open-source models despite U.S. export controls limiting chips. DeepSeek’s V3/R1 series previously sent shockwaves through Silicon Valley with its “Apache 2.0 license + performance benchmarks versus GPT-4 / Claude Opus,” and it had also been publicly accused by Anthropic earlier of “bypassing chip export controls through distillation.” This $20 billion valuation will also help DeepSeek move toward positioning itself as a “global top AI laboratory,” creating clear competitive pressure on leading U.S. frontier model vendors.

This article—DeepSeek in talks for its first round of external fundraising with a $20 billion valuation: a new high for Chinese AI valuations—first appeared on Lian News ABMedia.