News Summary

Meta’s Llama large language model (LLM) has surpassed 650 million downloads globally in 2024, marking its significant impact as a leading AI framework. The Llama’s journey included the release of Llama 3 and the groundbreaking Llama 3.1 in July, which featured the first frontier-level open AI system, the 405B model. At Connect 2024, Meta introduced Llama 3.2, its first-ever multimodal models optimized for edge devices and mobile platforms. The success of Llama is attributed to its extensive network of collaborators, such as AWS, AMD, Microsoft Azure, Dell, NVIDIA, Google Cloud, and Oracle Cloud, ensuring efficient operations across various environments. The AI model has seen accelerated adoption globally, particularly in Latin America, Europe, and the Asia-Pacific region. License approvals for Llama have doubled in the last six months, reinforcing its role as a universal AI model for developers worldwide. The open-source community has also played a crucial role, with over 85,000 Llama-derived models published on Hugging Face. Since its launch in February 2023, Llama has averaged one million downloads daily, with nearly 600 million active users of Meta AI tools globally.