China’s Big Fund is set to spearhead a $45 billion investment round in DeepSeek, a Hangzhou-based AI startup, as confirmed by four informed sources. This valuation, up from $20 billion just weeks prior, reflects investor confidence in DeepSeek’s open-source R1 model, which rivals U.S. AI giants like OpenAI despite using far less computing power. Key backers include Tencent and potentially founder Liang Wenfeng, who owns 89.5% of the firm via personal holdings.

The Big Fund, China’s state-backed semiconductor investment vehicle, aims to bolster domestic AI infrastructure amid U.S. export restrictions. With $47 billion raised in its third funding phase, it has previously supported SMIC and Yangtze Memory Technologies, positioning China to reduce reliance on foreign tech. DeepSeek’s latest V4 model now runs on Huawei’s Ascend 950PR chips, signaling a strategic shift as Huawei overtakes Nvidia in the Chinese market.

Liang Wenfeng, who initially sought funding to retain talent, may now inject personal capital to expand DeepSeek’s computing resources. This aligns with Beijing’s broader goal of creating a self-sufficient AI ecosystem, combining models, chips, and software. Analysts warn this could threaten U.S. dominance, with Nvidia’s Jensen Huang calling it a “horrible outcome” for America.

While DeepSeek prioritizes research over commercialization, peers like Zhipu ($52B market cap) are already generating revenue. The investment underscores China’s push to close the AI gap, even as its chip production lags two generations behind U.S. standards.