Meta bellied up to the artificial intelligence (AI) bar today to announce that its newest large language model, Llama-3, is the “most capable,” and “best open source model” currently available.
The company’s statements surrounding the general availability of Llama-3 as well as a new standalone “Meta AI” portal, has the tech world abuzz with declarations that the current space leaders, Microsoft, OpenAI, Google, and Anthropic, finally have some stiff competition from the company formerly called Facebook.
A Meta blog post made no quibbles about the company’s position concerning where its Llama suite of LLMs now lie in the global AI model hierarchy:
“This next generation of Llama demonstrates state-of-the-art performance on a wide range of industry benchmarks and offers new capabilities, including improved reasoning. We believe these are the best open source models of their class, period.”
Analysts and pundits have long awaited the arrival of Llama-3, with many claiming that preview versions demonstrated better reasoning and performance than contemporary challengers GPT and Gemini, comparable offerings from OpenAI and Google respectively.
The large language model market continues to expand year over year with updates to the most popular models seemingly occurring at a steady clip. Current market leader OpenAI is rumored to be on the verge of launching its latest model — GPT-5 if it follows previous naming conventions.
According to a report from Business Insider, the next GPT model will be “materially better” than its predecessors, but these rumors have yet to be substantiated.
As Cointelegraph recently reported, the LLM market is heating up with Microsoft and Google reportedly trading $100 billion blows. Microsoft and OpenAI pledged $100 billion to build a new AI super data center while Google DeepMind CEO Demis Hassabis has let on that he believes Google will spend more.
While the costs continue to rise, the rewards could be even greater. At stake, according to both Google and Microsoft, is the development of the world’s first artificial general intelligence (AGI). An AGI system, hypothetically-speaking, would be capable of doing any task a human could do, given the proper resources.
