Meta Makes Multimodal Move with LLaMA 4 Launch
Meta Platforms has unveiled the next generation of its artificial intelligence model, introducing LLaMA 4, which includes two new versions named LLaMA 4 Scout and LLaMA 4 Maverick. According to a statement released by the company on Saturday, the models are built on advanced multimodal technology capable of analyzing and interpreting a variety of data types, including text, images, videos, and audio. The announcement was reported by Reuters.
Describing the release, Meta said this is their “most advanced model to date” and “best in class for multimodality.” Both versions have been released as open-source models, promoting accessibility and further research in AI development.
In addition to Scout and Maverick, Meta also previewed a more powerful model titled LLaMA 4 Behemoth, which the company touts as “one of the most intelligent language models in the world.”
Following the success of ChatGPT and the rapidly evolving AI landscape, Meta has announced its plan to invest $65 billion in AI infrastructure in 2025, signaling its ambition to lead in the competitive field of artificial intelligence.
