Meta unveils new Llama 3.3 AI model

Meta Platforms (NASDAQ:META) took the wraps off its latest large language model, known as Llama 3.3, on Friday.

Meta shares rose 2.5% in midday trading on Friday.

Llama 3.3 is a 70B parameter model that is able to deliver the same performance as the company’s 405B model, “but is easier & more cost-efficient to run,” Meta AI executive Ahmad Al-Dahle said in a post on X.

Al-Dahle also showed that the new Llama 3.3 can outperform several other AI models, including Google’s (GOOG) (GOOGL) Gemini 1.5 Pro, OpenAI’s GPT-4o and Amazon’s (AMZN) Nova Pro, which the tech giant announced at its AWS re:Invent conference this week.

Llama 3.3 is available for download already, Meta added.

The release of the new AI model comes just a few months after Meta released its Llama 3.2 model in September. That model has mini versions available, which allows them to fit on edge and mobile devices.

Meta is also hard at work on its upcoming Llama models, known as Llama 4. Chief Executive Mark Zuckerberg talked up the Llama 4 models on the company’s third-quarter earnings call and said they are “being trained on a cluster that is bigger than 100,000 H100s or bigger than anything that I’ve seen reported for what others are doing.”