The latest version of their open-source LLM offers two configurations: 8B and 70B, which stand out among competitors due to high scores on benchmark tests. The innovation of Llama 3 lies in the fact that it was trained on a dataset seven times larger than the previous version—15 trillion tokens and four times more code. It is expected that in the coming months, an even more powerful model with over 400 billion parameters will be released, competing with giants such as GPT-4. Llama 3 will be integrated into nearly all Meta platforms, including Facebook, Instagram, WhatsApp, Messenger, and on the new Meta.ai site through the Meta AI assistant. This is a significant step in Mark Zuckerberg’s strategy to attract a broad audience to cutting-edge AI technologies, providing top-tier AI models to more than 3 billion users worldwide.