News

Meta Unveils Llama 4 AI Series Featuring New Expert-Based Architecture
The Llama 4 series is the first to use a “mixture of experts (MoE) architecture,” where only a few parts of the neural ...
InfoWorld2d
What misleading Meta Llama 4 benchmark scores show enterprise leaders about evaluating AI performance claims
AI benchmarking is critical to determine performance, but results can be irrelevant to enterprise workflows; enterprise ...
Meta’s surprise Llama 4 drop exposes the gap between AI ambition and reality
"The vibes around llama 4 so far are decidedly mid ," independent AI researcher Simon Willison told Ars Technica. Willison ...
Llama 4 : 10 Million Context Window Fully Tested
Llama 4 Scoot is a compact yet highly capable model, boasting 17 billion active parameters and 16 experts. Its standout ...
Meta defends Llama 4 release against ‘reports of mixed quality,’ blames bugs
Llama 4 continues to spread to other inference providers, but it's safe to say the initial release has not been a slam dunk.
Meta exec denies the company artificially boosted Llama 4’s benchmark scores
A Meta exec has denied a rumor that the company trained its AI models to present well on benchmarks while concealing the ...
Explore more
LLaMA 4 Maverick Review : Strengths, Weaknesses, and Real-World Performance
LLaMA 4 Maverick excels in reasoning but struggles with coding. Discover its strengths, weaknesses, and future potential in ...
LLMs from Meta: The fourth Llama herd has arrived
Llama 4 was released in a rush. After the enthusiasm generated by Llama 3, the sequel surprises and disappoints in some areas ...