During the festive reunion time of the Chinese New Year, an AI platform developed within China has captivated immense attention both domestically and internationally.

Before the festival, DeepSeek emerged dramatically, shaking the AI industry with its outstanding performance and remarkably cost-effective development expenses.

On January 27, US market time, stock prices plummeted sharply as trading opened, especially within the tech sectorAmong the most significant losses, American chip giant NVIDIA saw its shares drop approximately 17%, while the stock prices of Broadcom and AMD fell by 17% and 6%, respectivelyMicrosoft also faced a 2% declineFurthermore, AI-related derivatives, including power suppliers, were heavily impacted, leading to a 21% decrease for United Energy and a staggering 29% drop for Vistra.

The creation of DeepSeek has been a significant influence not only on the US stock market but also within domestic markets

Advertisements

Related enterprise Daily Interactive recorded its upper limit on the last trading day before Spring Festival on January 27, achieving a daily increase of 20%. Prior to this daily surge, the company’s stock had witnessed nearly a 20% rise in just one week.

In light of DeepSeek's favorable effects, the AIGC concept index in Hong Kong stocks has consistently risenAccording to data from iFinD, this index has accumulated a 27.23% increase over eight trading days starting from January 21.

The influence of DeepSeek did not cease with the pre-festival surgeDuring the Spring Festival, DeepSeek not only went viral with the topic of “Million Dollar Salary Recruitments” but also garnered the support of vast numbers of domestic and international users in the face of malicious attacks.

According to research reports from international investment banks, DeepSeek's valuation has now reached a staggering 10 billion USD

Advertisements

Notably, the changes brought by DeepSeek are not limited to just the AI industry.

“A Mysterious Force from the East”

DeepSeek, officially known as Hangzhou DeepSeek Artificial Intelligence Basic Technology Research Co., Ltd., was established on July 17, 2023, with a focus on developing advanced large language models (LLM) and associated technologiesThis startup technology company, born in China, has produced an open-source model comparable to OpenAI by virtue of its “refined” startup team.

According to public information, DeepSeek's founder Liang Wenfeng, a post-85s generation individual, became a “top score student” by achieving high marks in the Chinese college entrance examination at the age of 17, enrolling in Zhejiang University to study Electronic Information Engineering

Advertisements

In 2007, he entered graduate school at Zhejiang University, focusing on Information and Communication EngineeringIn 2013, Liang, along with his university classmate Xu Jin, founded Hangzhou Yakebi Investment Management Co., and two years later, they established Hangzhou Huanfang Technology Co., committed to quantitative investment through mathematics and artificial intelligence.

By 2021, Huanfang's asset management scale surpassed the trillion RMB mark, and in 2023, he announced a significant foray into the general artificial intelligence landscape with the establishment of DeepSeek, focusing on developing genuinely human-level AI.

By the end of 2024, DeepSeek plans to release its next-generation large language model, V3, while simultaneously announcing it as open-source

• Investing in A-Shares the Bogle Way
• Warning: The U.S. May Face Inflation
• UK Treasury Gold Sold at Discount
• Longi Green Energy: Profit Plunges to $8 Billion Loss
• Dow Jones Rises Consecutively

Test results indicate that it has outperformed several mainstream open-source models in various evaluation tests, also indicating cost advantages.

On January 20, 2025, coinciding with the opening day of the World Economic Forum 2025 annual meeting, DeepSeek unveiled its latest open-source model, R1, sparking global attention within the field of artificial intelligence again.

The company revealed that the R1 model represented a significant technological breakthrough—leveraging pure deep learning methods to enable AI to spontaneously develop reasoning capabilities, rivaling OpenAI's o1 model in tasks such as mathematics, coding, and natural language reasoning, all while maintaining the company's high cost-effectiveness advantage.

Reportedly, DeepSeek's R1 model training cost is approximately 5.6 million USD, significantly lower than the hundreds of millions or even billions of dollars invested by American tech giants like OpenAI, Google, and Meta in AI technologies.

In Silicon Valley, DeepSeek has early on been dubbed as the “Mystical Force from the East.” The release of the R1 model has drastically impacted the entire AI industry, momentarily affecting the stock prices of tech stocks in the US, and ranking at the top of global download charts on Apple App Store and Google Play Store for several consecutive days.

DeepSeek-R1 is recognized as one of the most advanced large language models available today, capable of delivering high-quality language processing abilities

In simple terms, DeepSeek has adopted even “smarter” algorithms that reduce AI training costs by nearly 60%, achieving performance that matches or surpasses similar modelsThis unique cost-performance ratio directly addresses a significant pain point in the industry.

Liu Wei, the director of the Human-Computer Interaction and Cognitive Engineering Lab at Beijing University of Posts and Telecommunications, noted that DeepSeek’s greatest advantage lies in the improvement and optimization of algorithms that save computational power and data volumeWhere OpenAI was “powerfully miraculous,” DeepSeek is proving that “small force can also achieve miracles” — employing lesser computational power to create innovative results using new methods.

Analysts predict that this development breaks the long-standing path dependency of the global AI industry on “computational piling,” with the DeepSeek model expected to inspire a wave of innovation, further advancing AI technology globally.

Huatai Securities commented that DeepSeek's low-cost, high-performance advantages have sparked discussions about capital expenses and application scenarios within the industry, potentially prompting investors to reevaluate the technological potential of Chinese tech corporations.

Zheshang Securities also expressed that the disruptive nature of DeepSeek-R1 represents the first successful attempt for a Chinese model to catch up to overseas counterparts, especially the globally leading AI models represented by OpenAI, realizing algorithmic optimization to achieve equity in AI and breaking the established pattern of achieving miracles through heavy computational power and chip advancements

This will lead to a reassessment of AI asset values worldwide.Zheshang Securities views DeepSeek as stimulating a global wave of AI innovation, propelling the continuous advancement of AI technologies.

Recently, numerous platforms, including Baidu Intelligent Cloud, Huawei Cloud, Alibaba Cloud, Tencent Cloud, 360 Digital Security, and Cloud Axis Technology, have announced the launch of the DeepSeek large modelInternationally, major tech firms like Amazon AWS, Microsoft Azure, and NVIDIA have likewise begun integrating the DeepSeek model.

The impact of DeepSeek on the AI industry is undoubtedly profound and pervasive.

Exploring Commercial Pathways

The impact of DeepSeek on the entire industry chain continues to unfold.

On January 27, DeepSeek created a buzz in the A-share market, with multiple stocks related to DeepSeek experiencing upward movement on that trading day

Data from Wind indicated that the Wind DeepSeek Cooperation Index surged over 13% during early trading, concluding the day with an ascent of 10.96%.

Among them, Daily Interactive and Zhaochuang Information closed at the 20% upper limit, while stocks like Tuoer Si, Parallel Technology, and Feilixin enjoyed substantial increasesZhejian Dongfang, who participated in DeepSeek’s angel funding round, saw a 38.2% increase in stock price over the past ten days; Huajin Capital, having participated in the Pre-A funding round for DeepSeek, observed a 21.7% increase in stock price within ten days.

Many companies have embraced this wave of “untold riches.” On February 4, Daily Interactive announced that they had connected to DeepSeek several stages ahead, completing evaluation and utilizing private deployment with their computational capabilities, stating that they would embrace the open technology ecosystem that the DeepSeek large model presents as the industry enters the era of large models.

DeepSeek has become a “hot topic,” with many companies capitalizing on this, while others have been eager to clarify

Zhaochuang Information issued a rebuttal, confirming the absence of established business cooperation with DeepSeek.

Despite claims of no collaborations from listed companies, investor curiosity and enthusiasm towards DeepSeek and its relevant industry chain remains unflaggingThe news regarding companies associated with DeepSeek continues to evolve, with numerous investors querying other listed companies about their use of DeepSeek’s large model.

Undeniably, DeepSeek’s success has ignited hope among many enterprisesInvestor Zhejiang Dongfang has forecast a net profit of 947 million RMB for 2024, representing over a 75% year-on-year increaseHuajin Capital is also exhibiting robust growth, anticipating profitability between 145 million to 185 million RMB, marking an expected increase of over 95% year-on-year.

However, within this sphere of “DeepSeek concept stocks,” several companies have issued profit loss warnings

Daily Interactive has forecasted a further net loss between 61.5 million and 91 million RMB for 2024, while projected net profit, excluding non-recurring items, is expected to be between 16 million to 23.5 million RMB.

The future commercialization of DeepSeek merits attentionGiven its stance on “open source,” industry watchers are keen to discover whether it can expand into advanced services to diversify its business portfolioIn the long-term, how DeepSeek can leverage its low-cost advantage to tap into critical technologies and large-scale computational capabilities remains a pressing question.

However, there is no doubt that the emergence of DeepSeek has disrupted the monopoly of traditional AI giants, reshaping the competitive landscape of the global AI market.