DeepSeek reportedly eyes $1.5B raise before IPO

DeepSeek, the Chinese large language model developer, is reportedly preparing for a 2027 IPO debut, with a listing that could arrive as early as the end of this year. The company is also said to be looking to raise around $1.5 billion in new funds at about a $71 billion valuation. The reported talks would follow a $7 billion funding round completed just a month ago at around a $50 billion valuation, described as its first-ever outside funding round.

Founded in 2023, DeepSeek drew attention after releasing AI technology described as more efficient and more cost-effective than U.S. model makers. Its popularity has grown quickly: in June, it accounted for nearly 23% of all the tens of trillions of tokens processed by enterprise-focused AI gateway Vercel, according to the company. Anthropic accounted for 32% of tokens in the same comparison.

TechCrunch frames DeepSeek as another sign that Chinese open source models can perform close to top U.S. AI labs despite U.S. export controls on chips. DeepSeek’s cloud service runs on chips made by Huawei Technologies, according to the article. Reported investors include Tencent and Beijing’s National Artificial Intelligence Industry Investment Fund, per Bloomberg.

The key questions are whether DeepSeek completes the reported fundraise, whether the valuation holds, and whether the IPO timeline lands in 2027 or accelerates. Readers should also watch how enterprise usage, chip access, and open source model performance shape investor appetite. DeepSeek could not be reached for comment, according to TechCrunch.