DeepSeek Closes $7 Billion Funding Round to Build Domestic AI Chip Infrastructure

DeepSeek has completed a $7 billion funding round, according to Pandaily, marking the largest venture capital raise by a Chinese AI laboratory. The proceeds will reportedly fund domestic AI chip development and procurement, a significant expansion of DeepSeek’s research staff, and accelerated work toward artificial general intelligence.

The Round

The close follows months of reported fundraising activity. In May, TechCrunch reported that DeepSeek was in talks to raise its first-ever outside investment at a valuation that had jumped from $20 billion to $45 billion in a matter of weeks. The Financial Times and Bloomberg, cited in the TechCrunch report, identified the state-backed China Integrated Circuit Industry Investment Fund as the round’s lead investor, with Tencent and Alibaba also reportedly in discussions to participate.

The $7 billion represents DeepSeek’s first venture round. Founder Liang Wenfeng, a hedge fund billionaire who controls approximately 90% of the company, had not previously sought outside investors. The fundraise was driven in part by the need to offer equity to employees after competitors began poaching DeepSeek’s researchers, according to TechCrunch.

Domestic Compute as Strategic Priority

The funding’s focus on domestic chip infrastructure reflects a broader Chinese strategy to reduce dependence on US semiconductor technology. DeepSeek has optimized its models to run on chips manufactured by Huawei Technologies, a combination Beijing views as critical to building sovereign AI capability.

The timing is significant. US export controls on frontier AI models, including the 18-day suspension of Anthropic’s Fable 5 and Mythos 5 that was lifted just this week, have created uncertainty for global enterprises relying on American AI providers. Chinese labs with independent compute infrastructure face no such restrictions on their own deployments.

Scale in Context

The $7 billion raise places DeepSeek among the most heavily funded AI companies globally. For comparison, Groq closed $647 million in growth funding last week, and OpenAI’s $122 billion round closed in March. DeepSeek’s raise is the largest single round by a Chinese AI startup.

DeepSeek’s models already power cost-sensitive agent deployments globally. Coinbase CEO Brian Armstrong disclosed in late June that routing inference to Chinese models cut AI infrastructure costs by 50%, according to The Decoder. The capital injection into domestic compute capacity could accelerate this cost advantage by securing dedicated chip supply for DeepSeek’s training and inference operations.