
On July 23, 2024, Meta introduced Llama 3.1, a suite of open-source artificial intelligence (AI) models that are poised to revolutionize the field of natural language processing (NLP). The release includes the Llama 3.1 405B model, which Meta claims is the world's largest and most capable openly available foundation model, along with upgraded versions of the 8B and 70B models.
According to Meta CEO Mark Zuckerberg, the latest generation of Llama will ignite new applications and modeling paradigms, including synthetic data generation to enable the improvement and training of smaller models, as well as model distillation—a capability that has never been achieved at this scale in open source.
The Llama 3.1 models boast state-of-the-art capabilities in general knowledge, steerability, math, tool use, and multilingual translation. The 405B model, in particular, rivals the top AI models and outperforms GPT-4o and Anthropic's Claude 3.5 Sonnet on several benchmarks. Meta trained the model using an extensive dataset comprising 15 trillion tokens, with information current up to 2024.

One of the key advantages of Llama 3.1 is its open-source nature. Unlike closed models, Llama model weights are available for download, allowing developers to fully customize the models for their needs and applications, train on new datasets, and conduct additional fine-tuning. This enables the broader developer community to more fully realize the power of generative AI and deploy the technology in any environment, including on-premises, in the cloud, or even locally on a laptop—all without sharing data with Meta.
Meta has also made changes to its license, allowing developers to use the outputs from Llama models—including the 405B—to improve other models. This could have important implications for how models work together and the economic implications for the return on investment of smaller models.
The Llama 3.1 models are available for download on llama.meta.com and Hugging Face, as well as for immediate development on Meta's broad ecosystem of partner platforms, including major cloud services like AWS, Azure, and Google. Companies such as Scale.AI, Dell, Deloitte, and others are ready to help enterprises adopt Llama and train custom models with their own data.
Alongside the release of Llama 3.1, Meta is expanding the reach of its Llama-powered AI assistant to additional countries and languages, while also introducing a new feature that enables image generation based on a person's unique appearance. The Meta AI assistant is now available in 22 countries, with users able to interact across WhatsApp, Instagram, Messenger, and Facebook in new languages, including French, German, Hindi, Italian, Portuguese, and Spanish.
The community has already built impressive applications with past Llama models, such as an AI study buddy deployed in WhatsApp and Messenger, an LLM tailored to the medical field, and a healthcare non-profit startup in Brazil that streamlines patient information communication. With the release of Llama 3.1, Meta anticipates a surge in innovation and the development of more advanced chatbots, tools with greater reasoning capabilities, and better computer coding agents.
As the AI race continues to heat up, Meta's commitment to open-source development sets it apart from the growing trend of secrecy among AI companies. While some developers have criticized Meta for not being fully transparent about certain aspects of its models, such as the training data, the release of Llama 3.1 represents a significant step forward in the democratization of AI technology.