China's DeepSeek reveales updated AI model as OpenAI rivalry intensifies - Reuters

Investing.com - DeepSeek has unveiled an update to its low-cost artificial intelligence model, intensifying the Chinese start-up's bid to rival AI industry leaders like ChatGPT-maker OpenAI and Anthropic, Reuters reported on Tuesday.

The newest version of the offering, DeepSeek-V3-0324, has become available through the AI development platform Hugging Face, the news agency said.

Reuters added that benchmark metrics on Hugging Face show that the model has displayed significant improvements in key areas like reasoning and coding when compared to its predecessor.

DeepSeek, which launched its V3 model in December and R1 offering in January, has emerged as a provider of competitive alternatives to Western AI counterparts at a fraction of the cost.

Silicon Valley executives and U.S. tech firm engineers have praised the V3 and R1 for delivering performance on par with OpenAI and Facebook-owner Meta Platforms' (NASDAQ: META) most advanced models. Depending on the task, the R1 in particular costs between 20 to 50 times less to use than the OpenAI o1 model, DeepSeek has said.

Although these claims have faced skepticism from some sections of the global AI landscape, DeepSeek's rise has still fueled worries among investors over the necessity -- and eventual financial returns -- of massive AI spending by mega-cap technology companies. Earlier this year, these fears sparked a sharp downturn in stocks, with AI chipmaker Nvidia (NASDAQ: NVDA) shedding $593 billion in market value on January 27, a record one-day decline for any company.

(Reuters contributed reporting.)

You May Also Be Interested In

• "Big Short" Steve Eisman says China’s AI models are "Achilles’ heel" of AI trade
• Micron: New Street upgrades to buy, says memory is no longer a cyclical story
• Software M&A heats up: Which SaaS stocks could be next for private equity?