Meta has launched Llama 4, its newest collection of AI models which now power the Meta AI assistant to provide enhanced experience.
The American-based tech giant on Saturday, April 5, 2025, revealed that LIama 4 can power Meta AI assistant on the web and in WhatsApp, Messenger, and Instagram.
The two new models are the “Llama 4 Scout” — a small model capable of “fitting in a single Nvidia H100 GPU” — and the “Llama 4 Maverick,” which is more akin to GPT-4o and Gemini 2.0 Flash.
Meta noted that it is still in the process of training Llama 4 Behemoth, which Meta CEO Mark Zuckerberg says is “the highest performing base model in the world.”
Related: Meta to build data centre in US state to support AI growth
Llama 4 Scout has a 10-million-token context window — the working memory of an AI model — and beats Google’s Gemma 3 and Gemini 2.0 Flash-Lite models, as well as the open-source Mistral 3.1, “across a broad range of widely reported benchmarks,” while still “fitting in a single Nvidia H100 GPU.”
For Llama 4, Meta claims it has switched to a “mixture of experts” (MoE) architecture, an approach that conserves resources by using only the parts of a model needed for a given task.
In a blog post, Meta stated, “These Llama 4 models mark the beginning of a new era for the Llama ecosystem.”
“This is just the beginning for the Llama 4 collection,” Meta added.
To note, Meta plans to discuss future plans for its AI models and products at the LlamaCon conference, which is slated to happen on April 29, 2025.
Related: Meta announces AI-powered ad tools to enhance e-commerce advertising
