The release is expected very soon, said a report by Financial Times citing a person familiar with the plans. "The new version of its open-source large language model can be used to build custom software and will be released 'imminently", said the report.
Earlier in February, Meta released its own large language model, known as LLaMA, to researchers and academics, but the new version is said to be more widely available and customisable by companies.
Meta' s LLMs are said to be “open-source,” which means that information about the new model will be made public. This will give an edge to Meta over OpenAI's GPT-4, which is kind of a "walled garden" meaning the data and code used to build the model are not accessible publicly or to third parties.
Notably, Meta has been working on AI research and development for more than a decade but is behind OpenAI’s ChatGPT and Google's Bard in the race of flagship generative AI models.
The upcoming software, which can create text, images and code, is however questionable when it comes to carbon footprint it would create, as it is powered by large language models (LLMs) which are trained on huge amounts of data and require vast computing power which in turn requires vast amount of energy to cool down the heat it will generate.
Advertisements