Meta’s open-source language model, Llama AI, is growing faster than ever. At the first-ever LlamaCon developer event, the company announced that downloads of its Llama models have reached 1.2 billion. That’s a major leap from the 1 billion mark it reported just a few weeks ago in March 2025—and nearly double the 650 million recorded last December.
Chris Cox, Meta’s Chief Product Officer, told developers at the event that the growth isn’t just about downloads. According to him, thousands of developers are now building tools on top of Llama. In fact, they’ve created tens of thousands of derivative models, many of which are downloaded hundreds of thousands of times each month.
It’s clear that Meta’s Llama ecosystem isn’t slowing down. Part of that success comes from its open nature. Unlike closed systems from companies like OpenAI or Anthropic, Llama models can be freely downloaded, customized, and deployed. This flexibility is especially appealing to startups, researchers, and companies looking to build AI into their own products without strict usage limits.
Meanwhile, Meta AI—the assistant powered by Llama—has reached around one billion users. It’s already embedded across Facebook, Instagram, Messenger, and WhatsApp. This gives Meta a massive advantage, offering real-world data at scale and constant feedback to improve its models.
Still, the competition is fierce. Just a day before LlamaCon, Alibaba released its latest Qwen3 model family, which delivers strong performance across popular AI benchmarks. Other major players like Google, Mistral, and Cohere are also pushing hard to lead in the open model race.
But Meta isn’t backing down. Its early move to open-source large language models gave it a head start. The company has also built a strong community around Llama AI, with active contributors and wide adoption across industries. This momentum makes it one of the most downloaded and widely used AI models today.
Looking ahead, Meta plans to invest more in safety tools and better documentation. As more people use Llama models, the risk of misuse grows. So Meta says it will roll out new guardrails to help developers build responsible, trustworthy AI apps.
For now, Llama AI is one of the fastest-growing ecosystems in the open-source AI world—and it shows no signs of slowing down.
