Meta debuts new Llama 4 models, but most powerful AI model is still to come

Meta on Saturday released the first models from its latest open-source artificial intelligence software Llama 4, as the company scrambles to lead the race to invest in generative AI. 

But the Facebook owner said it has not yet released the biggest and most powerful Llama 4 model, which outperforms other AI models in its class and serves as “a teacher for our new models.” That so-called Llama 4 Behemoth model is still in training, according to the company’s blog post. 

Llama 4 will help power AI agents, which will be capable of new levels of reasoning and action, Meta chief product officer Chris Cox said in March. Those agents will be able to surf the web and handle many tasks that could be of use to consumers and businesses.

Users can try two newly released Llama 4 models on Meta AI in WhatsApp, Messenger, Instagram Direct or the Meta AI website. That includes Llama 4 Scout and Llama 4 Maverick. 

“Our goal is to build the world’s leading AI, open source it, and make it universally accessible so that everyone in the world benefits,” Meta CEO Mark Zuckerberg said in a video on Instagtam. 

“And I’ve said for a while that I think that open source AI is going to become the leading models, and with Llama 4 this is starting to happen,” he said. “Meta AI is getting a big upgrade today.” 

Meta will host its first LlamaCon AI conference on April 29. The company is also expected to announce a standalone app for its Meta AI chatbot in the second quarter, CNBC reported in February.