• by Colin Kirkland , August 30, 2024

Meta launched its highly anticipated Llama 3.1 AI model just over a month ago. Now, the tech giant is set on letting the public know how its large language models are performing and how various companies that have integrated Llama into their day-to-day operations are utilizing the evolving technology.

In its new update on Llama models, Meta says that they are “approaching 350 million downloads to date” -- 10 times more downloads Meta's Llama models amassed at the same time last year. Over 20 million of those downloads took place in the past month, directly after the company released Llama 3.1.

Created in partnership with computing-chip distributor Nvidia, Llama 3.1 marks Meta's biggest open-source AI model to date, keeping the company in direct competition with tech giants and startups like Google, Amazon, OpenAI and Anthropic.

Compared to the Llama 3 models, Llama 3.1 is much larger, and it has 405 billion parameters trained with over 16,000 of Nvidia's H100 GPU chips (one of which costs between $30K and $40K), which can carry out a variety of tasks including coding, answering basic math questions and summarizing documents in eight languages.

“Meta AI is on track to reach our goal of becoming the most used AI assistant in the world by the end of the year,” Meta CEO Mark Zuckerberg said in an Instagram post last month, adding that Llama 3.1 is “smarter, supports more languages,” has “better reasoning” and will now become available in more countries.

Zuckerberg's prediction may come true, according to these new numbers, which show Llama usage growing ten times between January and July this year for some of Meta’s largest cloud-service providers including Amazon Web Services, Microsoft Azure, Databricks, Dell, Google Cloud, Groq, Nvidia, IBM watsonx, Scale AI and Snowflake.

Meta sees its open-source strategy as a way to deliver a diverse AI ecosystem to developers, giving them more choice and capability than closed models. Zuckerberg has been vocal about wanting to make open-source models commonplace as the technology continues to grow.

As for companies currently utilizing Llama 3.1, Meta says AT&T uses the AI model to fine-tune customer searches, DoorDash uses the technology to make tasks easier for its software engineers, and Niantic uses Llama to create real-time reactions of the creators called Dots in its AR mobile game, “Peridot.”

In addition, Zoom says it powers its AI Companion meeting summarizing tool with Llama and other large language models.