Llama 4: Here's what we know about Meta's most advanced AI model yet

The most recent iteration of Meta Platforms' large language model (LLM), Llama, known as Llama 4 Scout and Llama 4 Maverick, was launched on Saturday along with the opening of a new tab. Llama is a multimodal AI system, according to Meta. Text, audio, video, and picture data may all be processed and integrated via multimodal systems, which can also translate material between different forms. In a statement, Meta described the Llama 4 Scout and Llama 4 Maverick as its "most advanced models yet" and "the best in their class for multimodality."

In an Instagram video, Meta CEO Mark Zuckerberg shared the company's bold AI ambitions: "Our goal is to build the world's leading AI, open source it, and make it universally accessible... I've said for a while that open-source AI will lead the way, and with Llama 4, we're starting to see that happen."

Llama 4 Maverick and Llama 4 Scout will be available as open source software, Meta said. In addition, it claimed to be showcasing Llama 4 Behemoth, describing it as "one of the smartest LLMs in the world and our most powerful yet to serve as a teacher for our new models."

Following the success of OpenAI's ChatGPT, which changed the industry landscape and sparked investment in machine learning, large technology companies have started making significant investments in artificial intelligence (AI) infrastructure.

According to a story published by The Information on Friday, Meta postponed the release of its most recent version of LLM because, throughout development, Llama 4 failed to achieve Meta's technical requirements, especially in math and reasoning skills.

According to the source, the firm was also worried that Llama 4 couldn't have human-like voice chats as well as OpenAI's models. Amid investor pressure on large tech companies to demonstrate returns on their investments, Meta intends to invest up to $65 billion this year to enhance its AI infrastructure.