DeepSeek’s Low-Cost R1 Model Reshapes Generative AI Landscape

China-based DeepSeek shook the world of generative artificial intelligence (GenAI) early this year with the launch of its R1 model, a low-cost yet high-performance solution that posed a direct challenge to the dominance of OpenAI and other major players in the sector.
Since late 2022, only a handful of AI assistants, including OpenAI’s ChatGPT, Anthropic’s Claude, and Google’s Gemini, had dominated the market, thanks to multi-billion-dollar investments in engineers, data centres, and cutting-edge AI chips.
However, DeepSeek’s introduction of the R1 model, which it claimed cost just $6 million, upended the status quo. Powered by less advanced chips, the R1 model sparked significant debate across the industry.
While specialists speculated that DeepSeek’s true costs might have exceeded its claims, the model’s debut fuelled discussions about the growing commoditisation of GenAI assistants, driven by rapid innovation and market dynamics.
“The first company to train models must expend lots of resources to get there,” said Angelo Zino, senior equity analyst at CFRA. “The second mover can get there cheaper and more quickly.”
At the HumanX AI conference in Las Vegas earlier this week, Thomas Wolf, co-founder of Hugging Face, acknowledged the decreasing costs of launching GenAI models, suggesting that it was becoming less important which model users chose. “I feel like we are moving to this multi-model world, which is a good thing,” Wolf said, noting the lukewarm reception to the latest version of ChatGPT.
In contrast, Kevin Weil, chief product officer at OpenAI, pushed back against the idea that all models were now equal. “That’s actually not true,” Weil stated. “The days of us having a 12-month lead are probably gone, but I think we have a three- to six-month lead, and that is really valuable.” Weil outlined OpenAI’s strategy to maintain its edge over its competitors, explaining that the company’s 400 million users provided it with a unique advantage in continuously improving its models through massive data traffic.
“OpenAI has the Google advantage of being the thing that’s in everybody’s minds,” said Fen Zhao, research director at Alpha Edison equity firm.
Jeff Seibert, chief of the AI start-up Digits, agreed that OpenAI would remain ahead in certain advanced use cases but predicted that the gap between the top players would eventually close. “For advanced use cases, yes, there will be a lot of advantages,” Seibert said. “But for a lot of stuff, it won’t matter as much.” Seibert advised entrepreneurs to design their technology in a way that allowed for flexibility in swapping GenAI models as the industry continued to evolve.
The recent improvements in chip usage and optimisation techniques had driven down the costs of designing large language models (LLMs) like ChatGPT and Gemini, with some LLMs adopting open-source approaches that had accelerated innovation by making the software free for all to modify and enhance.
According to Zino, the valuations of closed-model startups such as Anthropic and OpenAI had likely peaked as their “first-mover advantage dissipated.” In February, SoftBank, a Japanese investment titan, invested $40 billion into OpenAI, raising the company’s valuation to $300 billion—almost double what it had been valued at the previous year.
“If you’re burning a billion dollars a month, which I think OpenAI is, you have to keep raising money,” said Jai Das, managing director at Sapphire Ventures. “I