Musk disclosed on platform X that the pre-training phase for the Grok V7 foundational model for xAI has been successfully concluded last week. This state-of-the-art model inherently embodies multimodality, enabling it to directly process video and audio bitstreams with ease.