.png)
The introduction of Meta Platforms' main "Behemoth" AI model has been delayed owing to worries about its capabilities, according to the Wall Street Journal, citing sources.
According to the article, company developers are failing to dramatically expand the capabilities of their Behemoth large-language model, raising concerns among employees about whether advances over previous versions are big enough to justify public release.
Behemoth was initially intended for release in April to coincide with Meta's debut AI conference for developers, but an internal aim for the model's deployment was later moved back to June, according to the article. According to the source, it will now be delayed until September or later.
A request for comment from Reuters was not immediately responded to by the social media behemoth.
Meta said in April that it was showcasing Llama 4 Behemoth, which it described as "one of the smartest LLMs in the world and our most powerful yet to serve as a teacher for our new models".
That month, it debuted the newest versions of the LLM Llama, known as the Llama 4 Scout and Llama 4 Maverick.
