When DeepSeek announced its $7 billion raise, the Chinese AI scene paused. The lab, which stunned industry insiders last year with a low‑cost, high‑performance language model, has closed its first round of outside capital, bringing in roughly 50 billion yuan—about $7 billion—at a valuation between $52 billion and $59 billion, according to Reuters.

Founder and chief executive Liang Wenfeng poured 20 billion yuan of his own money into the deal, securing a controlling stake. The remaining 30 billion yuan came from a small group of domestic investors. Tencent led the round with an estimated 10 billion yuan, while battery manufacturer CATL contributed around 5 billion yuan. The structure keeps Liang at the helm, a notable choice for a company that has never taken external venture capital.

DeepSeek has been funded entirely from the balance sheet of High‑Flyer, the quantitative hedge fund Liang also founded. The new round is described as a carefully managed “crack” rather than an opening of the doors. By retaining a controlling stake, Liang accepts the costs of outside capital—dilution, expectations, and scrutiny—while preserving strategic independence.

The lab’s open‑weight models, first released in January 2025, were reported to match or exceed the performance of Western systems such as OpenAI’s GPT‑4 while costing a fraction of the training spend. According to the company’s claims, its V3 model was trained for US$6 million, compared with the US$100 million cost of GPT‑4 in 2023, and used roughly one‑tenth the computing power of Meta’s comparable model. The breakthrough sent shock waves through the industry, prompting a reassessment of the compute requirements for frontier models.

DeepSeek’s success has positioned it as a state‑favoured project. The company has faced travel curbs on its talent and has been celebrated as a national AI champion. The presence of CATL—a battery maker rather than a software firm—on the cap table signals the breadth of Chinese corporate interest in backing a national AI leader.

The valuation itself is a leap for a company that, a year and a half ago, was little known outside specialist circles. The $7 billion raise at a valuation up to $59 billion reflects the market’s willingness to pay for the disruption DeepSeek has delivered.

The deal structure and investor mix also hint at DeepSeek’s priorities. By keeping control, Liang signals a commitment to the lab’s research‑first identity and its stated goal of developing artificial general intelligence. The inclusion of Tencent and CATL provides access to resources and networks that could accelerate product development and commercialization.

DeepSeek’s models are released under free and open‑source licenses, primarily the MIT License, and the company has recruited researchers from top Chinese universities as well as from outside traditional computer science fields. The lab’s use of mixture‑of‑experts layers and training on export‑grade AI chips allowed it to build powerful models while operating under ongoing trade restrictions.

The funding round is the largest AI investment in China to date, according to reports. It also follows a period of heightened scrutiny of AI companies by regulators, and a growing focus on domestic AI capabilities amid geopolitical tensions.

In the coming months, DeepSeek is expected to launch its next‑generation model, DeepSeek‑V4, which researchers say could rival frontier models such as OpenAI’s GPT‑5.5 and Anthropic’s Claude. The company has not yet announced a launch date, but the funding will likely support the development and scaling of the new model.

At present, DeepSeek remains a privately held company. The latest round has not yet been filed with the relevant securities regulators, and the precise allocation of shares may shift. The company has not issued a formal statement, but insiders say the deal structure reflects a deliberate strategy to balance external capital with founder control.

The $7 billion raise, the inclusion of Tencent and CATL, and the founder’s dominant stake together paint a picture of a Chinese AI lab that has moved from a niche research operation to a major player with significant domestic backing. The next steps for DeepSeek will be closely watched by investors, competitors, and regulators alike.