MUMBAI: DeepSeek is looking to put more yuan behind its artificial intelligence ambitions, raising approximately 100 billion yuan ($15 billion) in its latest funding round as it prepares to take on US AI heavyweights OpenAI and Anthropic with deeper pockets of its own.
The Hangzhou-based startup’s fundraising target reportedly doubled from its initial 50 billion yuan to 80 billion yuan before reaching around 100 billion yuan, according to Bloomberg News, which cited people familiar with the discussions. Chinese battery maker Contemporary Amperex Technology Co. Ltd. (CATL) and Tencent Holdings, the company behind WeChat, were among the biggest investors, the report said. The investors’ identities and the funding details have not been publicly disclosed.
The reported raise marks a significant shift for a company that made its name by challenging a central assumption of the AI race: that developing competitive models requires enormous spending. DeepSeek attracted global attention in 2025 after demonstrating that its models could rival leading Silicon Valley systems at substantially lower costs, unsettling markets and contributing to a sell-off that wiped approximately $1 trillion off global market values.
But building capable AI models on a leaner budget is one challenge; scaling them for a growing user base is another. DeepSeek has begun spending more aggressively on AI data centre infrastructure and developing additional open-source models for users in China and overseas. The fresh capital could help it expand that effort while competing in a market where computing power, research talent and infrastructure are increasingly expensive.
The funding would also help narrow the financial gap between DeepSeek and its American rivals, though it would not eliminate it. OpenAI and Anthropic have each raised more than $100 billion, according to Bloomberg’s report, giving them substantially greater financial firepower to develop models and expand their businesses.
DeepSeek’s rise has put pressure on competitors in China and beyond. Its approach of combining strong model performance with relatively low costs has attracted attention from customers, including in the US, while adding momentum to the push for open-source AI. Such models allow developers and businesses to adapt the underlying technology to their own requirements rather than relying entirely on proprietary services.
The company’s September launch of V4.1 Flash was followed by declines in the share prices of Chinese AI rivals MiniMax Group and Z.AI, according to the report. The episode underscored how quickly developments in the AI market can affect investor sentiment, particularly when a new model threatens to reset expectations around pricing and performance.
DeepSeek is not alone in pursuing a lower-cost approach. Chinese technology companies such as Alibaba Group Holding and ByteDance have also rolled out more affordable AI services, many of them based on open-source models. The competition is intensifying as companies seek to make AI more accessible while finding sustainable ways to fund the computing infrastructure and research needed to keep improving their systems.
Beyond expansion, DeepSeek is also reportedly preparing for a potential initial public offering (IPO). It is not the only Chinese AI player weighing a market debut, with rival Moonshot also working towards an IPO. Any eventual listings could offer investors a closer look at how China’s AI companies are valuing their technology, growth prospects and spending requirements.
Representatives of DeepSeek, Tencent and CATL did not respond to emailed requests for comment, Bloomberg reported. The reported funding terms have not been publicly confirmed by the companies.
For DeepSeek, the challenge now is to turn its reputation for doing more with less into a business capable of competing at scale. The latest reported raise suggests that even a cost-conscious challenger needs substantial capital to stay in the AI race.
