Meta Llama model downloads exceeded one billion, with amazing growth rate

Meta's open source artificial intelligence model family, Llama, has exceeded one billion downloads, an achievement announced by CEO Mark Zuckerberg on social media Threads. Compared with the 650 million downloads in December 2024, Llama's downloads surged by 153% in just three months.

Originally designed for researchers, Llama has become a popular choice for developers and AI enthusiasts around the world. Ahmed Al Dahler, vice president of Meta's Generative Artificial Intelligence Division, said on LinkedIn that Llama has evolved from a research project to the most widely used open source AI model.

Meta plans to release multiple Llama models in the coming months, including an "inference" model of OpenAI-like o3-mini and a multimodal-capable version. Zuckerberg also hinted that it may launch a model with "agent" function, with the ability to operate independently.

Last October, Zuckerberg confirmed that the Llama4 will be released in early 2025 and said that the smaller Llama4 model will be ready first. He mentioned that the training scale of Llama4 exceeds 100,000 H100 chips, far exceeding the scale currently known by other companies.

Llama has powered Meta's Facebook, Instagram, and WhatsApp and played a key role in the company's long-term AI strategy.