Llama 4 continues to spread to other inference providers, but it's safe to say the initial release has not been a slam dunk.

The Llama 4 series is the first to use a “mixture of experts (MoE) architecture,” where only a few parts of the neural ...

Llama is the most popular open-source model in the AI community, and the fourth generation arrived with big promises—and ...

Meta released Llama 4, its new collection of AI models on April 5, including Llama 4 Scout, Llama 4 Maverick, and Llama 4 ...

Explore the controversy around Meta's Llama 4, from transparency issues to performance benchmarks, and its impact on the AI industry.

The company was also previewing the Llama 4 Behemoth, described as their most powerful model so far, intended to guide and ...

Meta Platforms Inc. (NASDAQ: META) unveiled the first models of its latest open-source artificial intelligence (AI) software, ...

Meta Platforms on Saturday released two versions of Llama 4, the newest generation of its flagship open-source large language ...

This is a developing situation, and a surprising one for Meta, whose AI reputation has improved with each new Llama release before this one.

Meta's interpretation of our policy did not match what we expect from model providers." It also plans to upload the public ...

After someone on Threads asked why Llama 4 was released over the weekend, Meta CEO Mark Zuckerberg replied: “That’s when it was ready.” “It’s a very confusing release generally,” says ...

Meta attributed the mixed performance reports to implementation stability rather than flaws in the training process.