AI startup DeepSeek could raise $12 billion and increase its valuation by 1.5 times — Bloomberg

Chinese AI company DeepSeek is close to raising at least 80 billion yuan ($12 billion) in a new, second round of funding, according to Bloomberg sources. This significantly exceeds the startup’s own fundraising targets and will increase its valuation by nearly 1.5 times, according to calculations by Bloomberg Intelligence analysts. According to the agency, DeepSeek plans to go public early next year.

DeepSeek initially expected to raise about 50 billion yuan, but investor interest exceeded expectations following the successful release of the company’s latest AI model, V4 Flash, according to Bloomberg sources. Based on the preliminary documents already signed, the final investment amount could approach 100 billion yuan, the agency’s sources added.

The funding round will raise the company’s valuation to at least 500 billion yuan (about $75 billion), according to estimates by Bloomberg Intelligence analysts Robert Li and Jasmine Liu. In the previous round—in July—DeepSeek was valued at approximately 350 billion yuan.

According to Bloomberg, the largest investors will be leading battery manufacturer Contemporary Amperex Technology (CATL) and WeChat operator Tencent, the tech giant.

This funding lays the groundwork for one of the most eagerly anticipated debuts on the Chinese stock market in recent years, the agency notes. According to its sources, the next step after the round closes should be to restructure the company in preparation for its IPO.

Although discussions on the deal are nearing completion, the details could still change during the final negotiations, sources said. Representatives from CATL and Tencent declined to comment. DeepSeek did not respond to a request from Bloomberg.

DeepSeek is generating enormous interest among potential investors because it is one of the key players in China's strategy for global competition in the field of AI, the agency reports.

The company is on the rise following the release of its latest model, which has redefined the balance between cost and performance compared to market leaders Anthropic and OpenAI. The startup earned a reputation as a technological innovator after it first unveiled a model in 2025 that was comparable in capabilities to Silicon Valley’s developments but significantly cheaper.

“DeepSeek’s long-standing reluctance to seek outside funding is becoming a thing of the past (...) This change in course likely reflects the rising costs of developing AI models and processing tokens, which have forced the company to seek outside capital,” wrote analysts at Bloomberg Intelligence. However, they cautioned that it remains unclear how the startup will balance investors’ demands to maximize profits with the need to keep prices low to support Beijing’s plans for widespread AI adoption.

Like other players in the AI market, DeepSeek needs massive amounts of capital to develop its technologies. The company plans to deploy at least 160,000 state-of-the-art Huawei accelerators in its new large-scale data center in Inner Mongolia. This could result in one of the world’s largest clusters based on Huawei’s AI chips. In addition, DeepSeek has introduced open-source software for programming these chips, developed in collaboration with Huawei—an important step in tapping into a new niche in close partnership with a leading hardware manufacturer, the agency notes.

Moonshot AI, a competitor and developer of the popular Kimi K3 model, closed its latest private funding round on Tuesday, October 6, with a valuation of about $50 billion, and is preparing to hold an IPO in Hong Kong in the first quarter of 2027, according to Bloomberg sources. The company may begin preliminary meetings with investors in October and is considering raising up to $5 billion.

This article was AI-translated and verified by a human editor