DeepSeek is in talks to raise its first-ever venture round at a valuation that could reach $45B, more than double the $20B figure circulating just weeks ago. The round is reportedly led by the China Integrated Circuit Industry Investment Fund, the state vehicle Beijing uses to underwrite domestic chip and AI ambitions. Tencent and Alibaba are also said to be in discussions to participate.
The jump from $20B to $45B in a matter of weeks is the kind of repricing that usually requires a benchmark surprise or a product launch. Here it is something simpler: a Chinese AI lab that until now refused outside money is finally letting capital in, and demand has run well ahead of the paper.
DeepSeek has never taken venture funding. Founder Liang Wenfeng, a hedge fund manager who controls roughly 90% of the company, bankrolled the lab himself through his quant trading firm. The reason for the change, per the Financial Times, is mundane: rivals are poaching DeepSeek's researchers, and Liang needs equity to hand out.
Key facts
- 01DeepSeek is in talks for its first venture round at a potential $45B valuation, up from a $20B mark just weeks earlier.
- 02Founder Liang Wenfeng controls nearly 90% of the company and had never previously sought outside investors.
- 03The round is reportedly led by the state-backed China Integrated Circuit Industry Investment Fund, with Tencent and Alibaba in talks to join.
- 04DeepSeek rose to prominence in early 2025 with a model trained on a fraction of the compute and cost of U.S. rivals like OpenAI and Anthropic.
- 05The lab's models remain open weight on Hugging Face and are optimized to run on Huawei chips.
That detail matters more than the headline number. A lab that does not need money is being forced to the market by a talent war, which suggests Chinese AI compensation has crossed the threshold where prestige and mission alone do not retain staff. It is the same dynamic that pushed OpenAI and Anthropic into multi-hundred-billion-dollar valuations in part to fund retention grants.
“DeepSeek's potential valuation has more than doubled from $20B to $45B in a matter of weeks, before the lab has taken a single dollar of outside capital.”— Jaeden Schafer
DeepSeek became a global story in early 2025 when it released a large language model trained on a fraction of the compute and cost of leading U.S. systems. The lab has since kept reasonable pace with frontier models in reasoning and coding, and has done so while releasing weights openly on Hugging Face — a posture neither OpenAI nor Anthropic has matched at the frontier.
The strategic logic behind the round is not subtle. China is trying to build an AI stack that does not depend on U.S. chips, and DeepSeek's models have been optimized to run on hardware from Huawei Technologies. A well-capitalized DeepSeek paired with Huawei silicon is the closest thing the country has to a vertically integrated answer to the Nvidia-and-American-labs combination.
State backing through the Integrated Circuit fund makes the political framing explicit. This is not a purely commercial round; it is a policy instrument. Tencent and Alibaba joining would add cloud distribution and balance sheet depth, but the lead investor signals that DeepSeek's trajectory is now formally entwined with Beijing's industrial strategy.
There are reasons to be cautious about the $45B figure. It is a talks-stage number reported through secondary sources, not a closed round, and Chinese AI valuations have a history of moving on sentiment as much as fundamentals. DeepSeek did not respond to a request for comment, and neither the round size nor the precise structure has been disclosed.
The valuation also raises a harder question about open weights. DeepSeek's appeal to the global developer community rests on free model availability. Once state capital and strategic Chinese cloud investors hold meaningful equity, the pressure to monetize — through hosted APIs, enterprise deals, or restricted releases — grows. Liang has so far resisted that pull, but he will not own 90% forever.
For the rest of the AI market, a $45B DeepSeek changes the comparative math. U.S. labs have justified their valuations partly on the argument that frontier capability requires frontier spend. DeepSeek is the live counterexample, and a funded DeepSeek with employee equity, Huawei-tuned models, and Tencent and Alibaba distribution is a sharper version of that argument than the one that rattled American markets in early 2025.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
