China’s DeepSeek AI Challenges OpenAI – Triggers Historic Market Crash

China’s AI startup DeepSeek launched globally, entering Silicon Valley and Japan with a bang by unveiling its open-source model, DeepSeek-R1. This model delivers advanced AI capabilities at a fraction of the cost of industry giants like ChatGPT and Meta’s LIama. Although not entirely open-source, DeepSeek’s “open-weight” model allows public access to some aspects of its system, encouraging more developers to innovate. 

The release of DeepSeek triggered panic in the AI investment world, particularly in the US. Nvidia, the chipmaking giant and cornerstone of the AI revolution, suffered a historic $589 billion loss in market value, sparking widespread panic in the tech-heavy Nasdaq index and eliminating around $1 trillion from the global economy.

History of DeepSeek

DeepSeek, a Chinese artificial intelligence model based in Hangzhou, China, claims to be as powerful as OpenAI’s models but was developed in just a few months with a budget of under $6 million. DeepSeek’s AI model has stunned the global tech community. Liang Wenfeng, a graduate of Zhejiang University, founded DeepSeek in May 2023. Wenfeng also co-founded High-Flyer, a China-based quantitative hedge fund that owns DeepSeek.

Under the umbrella of High-Flyer, DeepSeek is an independent AI research lab that develops open-source LLMs. Its first model was released in November 2023. The company has iterated multiple times on its core LLM and created several variations. However, it didn’t become globally famous until January 2025, after the release of its R1 reasoning model.

DeepSeek Impact on the US Market

Nvidia’s stock declined sharply by 17%, resulting in billions of dollars in investors’ losses. CEO Jensen Huang alone lost $21 billion from his net worth. Another major player in AI chips, Broadcom, had its shares decline 17.4%, while Microsoft, Alphabet, and Meta faced declines ranging from 2% to 4%. The Nasdaq Composite fell 3%, dragging down the broader S&P 500 tech sector, which dropped 5.6%. 

Further, Power utility firms like Constellation Energy and Vistra, critical to the data centre infrastructure powering AI, saw their stocks fall by 20.8% and 28.3%, respectively. This was the sector’s worst one-day decline in over four years. By the market’s close, the global market had lost a remarkable $1 trillion value, and AI-driven stock gains reversed sharply.

DeepSeek Open-Source Approach

DeepSeek’s R1 code is publicly available under the MIT License. Developers can use, modify, and distribute the code without incurring licensing fees, enabling broader integration into various systems and applications. DeepSeek has become a significant game-changer in the AI industry, shaking the US tech industry. It was the most downloaded app on the iPhone App Store in the United States, a massive accomplishment for a Chinese-developed app. 

However, this led many to question how effectively some US tech companies, which have invested billions in AI development, have utilised their resources. What makes DeepSeek models unique is their performance and open-source nature with open weights, allowing anyone to build on top of them. The DeepSeek-V3 has been trained on a meagre $6 million, a fraction of the hundreds of millions invested in their frontier models by OpenAI, Meta, etc.

Also Read: LA Fires: How Climate Change Impacts the Economy

DeepSeek Vs OpenAI: Which One is Better?

Both DeepSeek and ChatGPT perform strongly in generating answers and solving complex problems. Developed by OpenAI, ChatGPT offers several features, including the ability to generate AI images, provide multimodal interfaces for tasks like image analysis, and interact with various tools such as Canvas. In addition, ChatGPT supports voice interaction, offering a more dynamic and flexible user experience, while DeepSeek is restricted to text-based searches.

DeepSeek’s model training costs 10% of OpenAI’s, yet it gives tough competition in answering queries. If DeepSeek’s claims hold, the belief that AI advancement requires increasing computational power would be disrupted. 

So, which one is better? Let’s see how they differ significantly in several ways. 

The US and China are driving the development of AI, but where does India stand in this high-stakes race? If Indian models struggle to keep pace, they risk falling behind. DeepSeek’s open-source approach has caught attention worldwide, enabling innovators to innovate and customise more. Experts have pointed out that DeepSeek’s success challenges the notion that you must invest considerably to get cutting-edge AI.