Chinese AI startup DeepSeek is considering raising external funding for the first time, a move that would mark a significant shift for a company that has so far relied on backing from its parent, hedge fund High-Flyer Capital Management, according to a report by The Information, citing five people familiar with the matter.
The startup is reportedly seeking at least US$300 million at a valuation of US$10 billion or more. The potential fundraising reflects rising capital requirements in the global race to develop frontier AI systems, where compute costs and talent retention have become increasingly competitive pressures.
DeepSeek has previously rejected funding offers from major Chinese venture capital firms and large technology companies, according to The Information. Founder and chief executive Liang Wenfeng has been described as favoring independence and minimal commercial influence over the company's direction.
However, sources cited by The Information said that DeepSeek has recently begun engaging with potential investors as it seeks additional resources to expand computing capacity and retain key researchers. The company has also experienced some talent turnover, with senior researchers reportedly joining competitors including Xiaomi and ByteDance.
DeepSeek has not released a major new model since the success of its R1 system in early 2025, and development of its next-generation model, V4, has reportedly been delayed by engineering complexity and hardware compatibility work.
Rising pressure from AI competition and infrastructure demands
Other industry reports suggest DeepSeek's fundraising considerations come amid broader shifts in China's AI sector in 2026, where competition has intensified, and access to compute infrastructure has become a critical constraint.
According to reporting compiled from multiple Chinese and international outlets, DeepSeek has been working to expand the capabilities of its upcoming models, including a significant increase in the context window size and extended coverage of training data. In February 2026, media reports indicated that the company had expanded its model context window from around 128,000 tokens to over 1 million, reflecting a push to handle longer, more complex inputs.
At the same time, DeepSeek has been adapting its systems to run more efficiently on domestic Chinese semiconductor platforms. Industry reporting in early 2026 suggested that the company has been prioritizing compatibility with Huawei chip architectures, a shift away from earlier reliance on Nvidia hardware due to export restrictions and supply-chain constraints.
This transition has reportedly contributed to delays in the rollout of its V4 model, as engineers work to optimize performance across different chip ecosystems.
Geopolitical scrutiny and shifting adoption patterns
DeepSeek's rise has also attracted international attention. A report from Associated Press in January 2026 noted that the company's AI systems have seen growing adoption in some developing markets, where lower-cost or more accessible AI tools are in demand. The same reporting also highlighted ongoing security concerns in several countries, as well as restrictions on internal usage by some Western organizations.
Meanwhile, other industry coverage in 2026 has pointed to heightened geopolitical sensitivity toward advanced AI systems developed in China, particularly as global governments tighten scrutiny of AI model capabilities, data use, and infrastructure dependencies.
Axios has also reported ongoing policy discussions in the US regarding the competitive positioning of Chinese AI firms, such as DeepSeek, particularly regarding supply chains and access to advanced semiconductors.
Article edited by Jack Wu
