Chinese Startup Disrupts US, Now Valued at Rp 866 Trillion
Jakarta, CNBC Indonesia - The Chinese artificial intelligence (AI) startup DeepSeek is reportedly preparing to launch its first funding round with a valuation reaching US$50 billion, equivalent to approximately Rp 866 trillion.
This information was revealed by three Reuters sources familiar with the company’s internal discussions. DeepSeek was previously known as an AI firm that rejected external funding.
DeepSeek shook the US technology industry when it first launched a low-cost AI model with performance capable of rivaling the expensive AI developed by US tech giants. That move triggered a global tech stock sell-off and once wiped out US$593 billion from Nvidia’s market value in a single day.
The Chinese government, through a national AI fund worth 60 billion yuan, is reportedly in discussions to become the lead investor in the funding round. Additionally, tech giant Tencent Holdings is said to be exploring an investment.
In this debut funding round, DeepSeek could potentially secure fresh capital of up to US$4 billion.
The funds will be used to expand AI computing capacity as well as improve facilities and employee welfare.
This expansion comes as DeepSeek faces pressure from domestic rivals such as ByteDance, Alibaba, and AI startups like MiniMax and Moonshot AI, which are aggressively spending to develop next-generation AI models.
The name DeepSeek made waves in the global technology industry after their open-source AI model went viral and sparked panic in the tech stock market last year.
At that time, DeepSeek’s affordable yet efficient AI model was seen as a serious threat to the dominance of US AI companies.
DeepSeek’s founder, Liang Wenfeng, has long been known as a figure reluctant to take his company public via IPO or submit to China’s major tech conglomerates. The company’s operations have largely been funded by his own hedge fund, High-Flyer.
However, the AI industry dynamics are changing rapidly. Development focus is now shifting from chatbots to AI agents capable of performing complex tasks automatically with minimal human intervention.
DeepSeek recently launched its latest AI model, V4, which is claimed to redefine open-source AI. Nevertheless, several independent evaluations indicate that its performance still trails behind the best AI models from both the United States and China.
The V4 launch also did not trigger a massive global tech stock sell-off like what happened last year when DeepSeek released the V3 and R1 models.