What is Chinese AI startup DeepSeek?

The latest artificial intelligence (AI) models launched by Chinese startup DeepSeek have spurred turmoil in the technology sector following its emergence as a potential rival to leading U.S.-based firms.

DeepSeek wrote in a paper last month that it trained its DeepSeek-V3 model with less than $6 million worth of computing power from what it says are 2,000 Nvidia H800 chips to achieve a level of performance on par with the most advanced models from OpenAI and Meta.

Those chips are less advanced than the most cutting edge chips on the market, which are subject to export controls, though DeepSeek claims it overcomes that disadvantage with innovative AI training techniques. DeepSeek’s AI assistant, which is powered by the DeepSeek-V3 model, surpassed OpenAI’s ChatGPT as the top-rated free application in the Apple App Store in the U.S.

The China-based firm’s emergence has raised questions about leading U.S. tech companies investing billions of dollars in advanced chips and large data centers used to train AI models. It also serves as a “Sputnik moment” for the AI race between the U.S. and China following the perception that the U.S. had an edge over its geopolitical rival in the emerging field.The quality of DeepSeek’s models and its reported cost efficiency have changed the narrative that China’s AI firms are trailing their U.S. counterparts, which began after the first Chinese ChatGPT equivalent was released by Baidu.

The DeepSeek-R1 model was released last week and is 20 to 50 times cheaper to use than OpenAI’s o1 model, depending on the task, according to a post on the company’s official WeChat account.

The R1 model is also open source and available to users for free, while OpenAI’s ChatGPT Pro Plan costs $200 per month.Getty Images

DeepSeek was created in late 2023 after controlling shareholder Liang Wenfeng, the co-founder of quantitative hedge fund High-Flyer, moved to create a “new and independent group, to explore the essence of [artificial general intelligence].”

Artificial general intelligence (AGI) is also a goal being pursued by OpenAI, which defines AGI as autonomous systems that surpass humans in most economically valuable tasks.

“DeepSeek did something really impressive with far fewer employees and funding than openAI did, and in less time,” said Mark Malek, CIO at Siebert. Malek noted that DeepSeek, “DOESN’T COMPETE WITH OPENAI,” and went to explain some of the differences between DeepSek and more well-known AI apps.

“OpenAI (and Google’s Gemini) is a broad, general-purpose tool, based on a vast corpus of information. It can be used to create more narrowly focused applications as well. OpenAI has vast capabilities in natural language processing, while DeepSeek is created to be task-specific,” Malek said. “Now, that doesn’t mean that DeepSeek is not good. What is important to understand is that it is not the same thing as OpenAI, so it would logically take fewer resources.”

Reuters contributed to this report.