New York – Meta’s Llama artificial intelligence models are being utilized by prominent companies such as Goldman Sachs and AT&T for various business functions, including customer service, document analysis, and code generation, according to a statement released by Meta on Thursday.
The Llama models, which are mostly available for free, have been downloaded nearly 350 million times since Meta first made them publicly accessible last year. This marks an increase from the 300 million downloads reported when Meta introduced the latest and largest version, Llama 3, in late July.
Meta also noted a significant rise in usage through cloud providers like Amazon Web Services and Microsoft Azure, with usage more than doubling between May and July of this year.
This announcement comes as Meta, alongside other tech giants heavily investing in AI, faces scrutiny from investors regarding the widespread adoption of AI technology and the potential return on these substantial investments.
Meta’s CEO, Mark Zuckerberg, has emphasized that offering cutting-edge AI models for free is a strategic move to prevent the company from being limited by a competitor’s proprietary technology. “The way for Llama to become an industry standard is by consistently being competitive, efficient, and open with each new generation,” Zuckerberg wrote last month.
While large language models like Llama have impressed users with their capability to produce human-like text, they still encounter challenges with certain logical tasks and are prone to factual inaccuracies, which can limit their utility in some business scenarios.
Despite these challenges, Meta highlighted that several major enterprises are experimenting with Llama models, showcasing their effectiveness and competitiveness compared to paid alternatives, such as those from leading AI provider OpenAI.
Other companies reported to be using Llama include Japanese financial institution Nomura Holdings, food delivery service DoorDash, and professional services firm Accenture.
