Meta Introduces Llama 2: The Next Generation ChatGPT Replacement

Meta and Microsoft, two prominent Big Tech companies, have joined forces to introduce Llama 2, an open-source large language model developed by Meta. Llama 2 is set to be integrated into Microsoft’s Windows operating system and its cloud computing platform Azure.

In an announcement made on July 18, the partnership was unveiled, revealing that Llama 2 would be freely accessible for both research and commercial purposes. Additionally, the model has been specifically optimized to operate seamlessly on Windows.

This collaboration substantiates previous speculations from the past week, which suggested that Llama 2 was designed to empower businesses and researchers in building applications on Meta’s AI technology infrastructure.

Meta claimed Llama 2 was trained on 40% more publicly available online data sources and can process twice as much context compared to Llama 1.

According to Meta’s statements, Llama 2 surpasses numerous competing open-source large language models (LLMs) in areas such as coding, proficiency, reasoning, and performance on knowledge tests. Nevertheless, Meta acknowledged that it falls slightly short in terms of efficiency compared to closed-source competitors like OpenAI’s GPT-4, as stated in one of their research papers.

In a July 18 Instagram post, Meta CEO Mark Zuckerberg said Llama 2 “gives researchers and businesses access to build with our next generation large language model as the foundation of their work.”

source