DeepSeek's Low Prices Undercut Big-Spending Competitors, Melius Research Says

DeepSeek, a Chinese AI startup, has sent shockwaves through the tech industry with its cost-effective AI models, challenging the dominant narrative of high capital expenditure in AI development. Melius Research has taken notice, highlighting the potential impact of DeepSeek's low prices on big-spending competitors and the broader AI industry.

DeepSeek's R1 model, released last week, has matched or even outperformed OpenAI's o1 model on several benchmarks, despite being trained at a fraction of the cost. The company spent just $5.6 million on rented GPU hours to train R1, compared to the hundreds of millions reportedly spent by U.S.-based leaders like OpenAI and Meta. Moreover, DeepSeek charges about one-thirtieth of the price it costs OpenAI's o1 to run, while still maintaining a small profit.

Melius Research analyst Tom Plunkett commented on the situation, stating, "DeepSeek's low prices are a wake-up call for big-spending AI companies. They need to reevaluate their strategies and consider more cost-effective approaches to AI development."

DeepSeek's success can be attributed to several factors, including its open-source approach, optimized hardware and data center, innovative software and algorithms, and access to affordable hardware. These factors have allowed the company to achieve comparable AI performance at a fraction of the cost compared to its U.S.-based counterparts.



The emergence of low-cost AI alternatives like DeepSeek could significantly impact the demand for high-performance GPUs and other computing resources. If AI companies can achieve similar results with less expensive hardware, this could lead to a decrease in demand for high-end GPUs, potentially impacting companies like Nvidia's sales growth and margins.

However, it is essential to consider that the AI landscape is rapidly evolving, and new technologies and approaches may emerge that could mitigate or even reverse these potential impacts. The broader AI industry may be influenced by DeepSeek's success in various ways, such as increased focus on software optimization, re-evaluation of hardware requirements, and greater interest in open-source AI development.

In conclusion, DeepSeek's low prices have undercut big-spending competitors, according to Melius Research. The company's cost-effective AI models challenge the dominant narrative of high capital expenditure in AI development and raise questions about the long-term competitiveness of U.S. tech companies. The broader AI industry may be influenced by DeepSeek's success, potentially leading to a shift in demand for high-performance GPUs and other computing resources. However, the ultimate impact on the AI industry and companies like Nvidia remains to be seen.