Elon Musk’s xAI company has integrated image understanding capabilities into its Grok model. Now, users with a paid subscription to the X social platform can upload images and ask the AI questions about their content. An xAI representative confirmed the new features in a post on X.
The new feature allows Grok to explain jokes in images, whenever someone needs it. Although this functionality is currently in beta, the developers promise further improvements.
Early users of the new feature are already reporting that the AI can’t always “understand” basic memes. Others mention difficulties uploading images—sometimes they have to upload them several times. Also, Grok still can’t generate images in response to analyzing other images.
This update aligns with the company’s strategy to create multimodal capabilities for its models. In August this year, Grok-2 was released with image generation support thanks to FLUX.1 technology.
The owner of the X social network, Elon Musk, hinted that upcoming updates will allow Grok to process documents in formats such as PDF. According to Musk, the team is achieving goals in just a few months that would take others years.
