Chinese artificial intelligence firm DeepSeek rocked markets this week with claims its new AI model outperforms OpenAI’s and cost a fraction of the price to build.

The assertions — specifically that DeepSeek’s large language model cost just $5.6 million to train — have sparked concerns over the eyewatering sums that tech giants are currently spending on computing infrastructure required to train and run advanced AI workloads.

But not everyone is convinced by DeepSeek’s claims.

CNBC asked industry experts for their views on DeepSeek, and how it actually compares to OpenAI, creator of viral chatbot ChatGPT which sparked the AI revolution.

Last week, DeepSeek released R1, its new reasoning model that rivals OpenAI’s o1. A reasoning model is a large language model that breaks prompts down into smaller pieces and considers multiple approaches before generating a response. It is designed to process complex problems in a similar way to humans.

DeepSeek was founded in 2023 by Liang Wenfeng, co-founder of AI-focused quantitative hedge fund High-Flyer, to focus on large language models and reaching artificial general intelligence, or AGI.

AGI as a concept loosely refers to the idea of an AI that equals or surpasses human intellect on a wide range of tasks.

Much of the technology behind R1 isn’t new. What is notable, however, is that DeepSeek is the first to deploy it in a high-performing AI model with — according to the company — considerable reductions in power requirements.

“The takeaway is that there are many possibilities to develop this industry. The high-end chip/capital intensive way is one technological approach,” said Xiaomeng Lu, director of Eurasia Group’s geo-technology practice.

“But DeepSeek proves we are still in the nascent stage of AI development and the path established by OpenAI may not be the only route to highly capable AI.” 

In terms of performance, DeepSeek says its R1 model achieves performance comparable to OpenAI’s o1 on reasoning tasks, citing benchmarks including AIME 2024, Codeforces, GPQA Diamond, MATH-500, MMLU and SWE-bench Verified.

In a technical report, the company said its V3 model had a training cost of only $5.6 million — a fraction of the billions of dollars that notable Western AI labs such as OpenAI and Anthropic have spent to train and run their foundational AI models. It isn’t yet clear how much DeepSeek costs to run, however.

If the training costs are accurate, though, it means the model was developed at a fraction of the cost of rival models by OpenAI, Anthropic, Google and others.

Daniel Newman, CEO of tech insight firm The Futurum Group, said these developments suggest “a massive breakthrough,” although he shed some doubt on the exact figures.

“I believe the breakthroughs of DeepSeek indicate a meaningful inflection for scaling laws and are a real necessity,” he said. “Having said that, there are still a lot of questions and uncertainties around the full picture of costs as it pertains to the development of DeepSeek.”

Meanwhile, Paul Triolio, senior VP for China and technology policy lead at advisory firm DGA Group, noted it was difficult to draw a direct comparison between DeepSeek’s model cost and that of major U.S. developers.

“The 5.6 million figure for DeepSeek V3 was just for one training run, and the company stressed that this did not represent the overall cost of R&D to develop the model,” he said. “The overall cost then was likely significantly higher, but still lower than the amount spent by major US AI companies.” 

DeepSeek wasn’t immediately available for comment when contacted by CNBC.

However, in comments to CNBC last week, Scale AI CEO Alexandr Wang, said he believed DeepSeek used the banned chips — a claim that DeepSeek denies.

Nvidia has since come out and said that the GPUs that DeepSeek used were fully export-compliant.

Yann LeCun, chief AI scientist at Meta, said that DeepSeek’s success represented a victory for open-source AI models, not necessarily a win for China over the U.S. Meta is behind a popular open-source AI model called Llama.

“To people who see the performance of DeepSeek and think: ‘China is surpassing the US in AI.’ You are reading this wrong. The correct reading is: ‘Open source models are surpassing proprietary ones’,” he said in a post on LinkedIn.

“DeepSeek has profited from open research and open source (e.g. PyTorch and Llama from Meta). They came up with new ideas and built them on top of other people’s work. Because their work is published and open source, everyone can profit from it. That is the power of open research and open source.”

WATCH: Why DeepSeek is putting America’s AI lead in jeopardy

– CNBC’s Katrina Bishop and Hayden Field contributed to this report

Meta CEO Mark Zuckerberg praised the Trump administration for backing Silicon Valley on a call with investors, adding that 2025 will be big for “redefining” the company’s relationships with governments.

“We now have a U.S. administration that is proud of our leading companies, prioritizes American technology winning and that will defend our values and interests abroad,” Zuckerberg said Wednesday. “I am optimistic about the progress and innovation that this can unlock, so this is going to be a big year.”

Meta on Wednesday also agreed to pay $25 million to settle a lawsuit with President Donald Trump, according to NBC News. Trump sued Meta after the company suspended his Facebook and Instagram accounts following the insurrection at the U.S. Capitol on Jan. 6, 2021.

Zuckerberg and Meta have made several public efforts to smooth over relations with President Donald Trump since his victory in November. The company donated $1 million to Trump’s inaugural fund late last year, weeks after Zuckerberg dined with him privately at his Mar-a-Lago resort.

Earlier this month, Zuckerberg announced that Meta would eliminate third-party fact-checking to “restore free expression” to the company’s platforms. He said the fact-checkers had been “too politically biased” and “destroyed more trust than they’ve created, especially in the U.S.”

The move was widely recognized as a nod to Trump, as he and other Republicans have long claimed that Meta’s platforms like Facebook and Instagram censor conservative views. Zuckerberg and Trump have had an especially rocky relationship in the past, as Trump has previously threatened the tech executive with life in prison.

The company also elevated Joel Kaplan, former White House deputy chief of staff under President George W. Bush with longstanding ties to the Republican Party, to its chief policy role earlier this month.

Zuckerberg’s public concessions appear to be earning him some good will, as he attended Trump’s inauguration alongside other tech moguls like Tesla CEO Elon Musk, Google CEO Sundar Pichai and Amazon founder Jeff Bezos this month.

Shares of Meta were up slightly in extended trading Wednesday after the company reported fourth-quarter earnings that beat Wall Street’s expectations on top and bottom lines.

–CNBC’s Jonathan Vanian contributed to this report

WATCH: Meta beats on top and bottom lines, stock slips on Q1 revenue guidance

Meta continues to lose billions of dollars developing the virtual reality and augmented reality technologies needed to underpin the nascent metaverse.

The social media giant reported fourth-quarter earnings Wednesday and said its Reality Labs unit recorded an operating loss of $4.97 billion while generating $1.1 billion in sales. Analysts were projecting that unit to log a fourth-quarter operating loss of $5.4 billion on $1.1 billion in sales.

Reality Labs is Meta’s unit that makes the Quest family of virtual-reality headsets and Ray-Ban Meta Smart Glasses.

Meta CEO Mark Zuckerberg kick-started his company’s VR endeavors in 2014 when it acquired the startup Oculus for $2 billion. Since then, Zuckerberg has characterized VR and AR as central to his plans to develop the futuristic digital world known as the metaverse, which he has said represent the next major computing platform.

Wall Street has questioned Zuckerberg’s metaverse investment. Reality Labs has tallied an operating loss of more than $60 billion since 2020, as of Meta’s fourth-quarter earnings report.

Meta last week said it would invest between $60 billion and $65 billion in 2025 capital expenditures to expand its computing infrastructure related to artificial intelligence. Zuckerberg has previously said AI is core to the company’s metaverse efforts, including its Ray-Ban Meta smart glasses. Meta develops that device with France-based EssilorLuxottica.

The social media company last year also unveiled its Orion prototype AR headset that is capable of overlaying digital objects on top of a person’s real field of view.

Meta released its latest VR headset, the $299 Quest 3S, during its September Connect event and pitched the device as a way for people to watch movies, play games and workout in VR.

Other tech companies are also investing in VR and AR.

Apple’s Vision Pro headset went on sale in the U.S. in February 2024 with a starting price of $3,499, and in December, Google and Samsung said they were working on a VR and AR device dubbed Project Moohan that will be available to buy in 2025 for an undisclosed price.