Made with Grok Imagine
According to Choi, following the update, Grok Imagine is no longer just an image generator but has become a more versatile tool for working with visual content. The model allows users to edit photos, combine up to five reference images, create images with text, resize images, and remove backgrounds.
What the Updated Grok Imagine Can Do
Min Choi demonstrated the capabilities of Grok Imagine 2.0 using ten examples, accompanied by the prompts he used.
These included Polaroid-style photos, photorealistic scenes that would be impossible to capture with a regular camera, collages, visual diagrams of cocktail recipes, and glossy magazine covers.
One example was a scene featuring a cheetah running across water. Other works demonstrate the ability to use multiple visual sources and create more complex compositions.
Choi paid particular attention to the ability to perform “surgical” photo editing—that is, making precise changes to an image without having to recreate it from scratch. Another announced feature is the ability to work with multiple references—up to five images at once.
The updated version also promises higher-quality text generation within images, resizing of finished images, and background removal.
Thus, Grok Imagine 2.0 shifts its focus from simply generating images based on text prompts to comprehensive work with visual content.
A New Phase in AI Model Competition
Choi’s demonstration highlights a trend in which image generators are gradually evolving into full-fledged tools for creating and editing visual content.
Users can not only describe the desired image in a prompt but also use existing images as a basis, combine multiple references, modify individual elements, and create content with a specific visual style.
Choi’s post, published on August 11, has received over 1.3 million views, according to the post’s metrics.
Musk’s brief comment, “Grok Imagine,” further drew attention to the update to xAI’s product and its capabilities in the field of generative graphics.
