- AI Creator Studio
- Posts
- Grok gets a major Image Editing update—but is it any good?
Grok gets a major Image Editing update—but is it any good?
Radu Lis
March 25, 2025
Hey there, creative human! 🌟
Earlier this week, X introduced a new "Edit Image" feature in Grok, allowing users to make adjustments to any image simply by describing them in a prompt.
In today’s newsletter, I’m taking you behind the scenes to explore Grok’s image editing capabilities and the real-world use cases that could transform your workflow. Let’s dive in and see what this new tool can really do! 👇🏻
STORY OF THE WEEK
🖼️ Grok 3 New AI Image Editor is here
From what's been presented, Grok 3 seems to perform well at transforming images into various styles — oil painting, Van Gogh style, Minecraft and so on.
And what better image to test with than the most famous wallpaper ever created?
 Original image |  “transform this image into the style of Leonardo da Vinci” |
 “transform this image into the style of Van Gogh” |  “change this image to a Minecraft style” |
Point proven. Grok can adjust the style remarkably well in no time, maintaining the essence of the original while convincingly applying distinctive artistic techniques.
Now I was interested in adding/removing objects from images, an area where Google Gemini 2.0 Flash has proven to excel:
 Original image |  “remove the notebook” |
 “replace the bottle of water with an exquisite 30-year-old Bordeaux wine” |  “add a 50-year-old German woman who is seated on a chair, facing the camera” |
When I prompted for the Bordeaux wine, most outputs rendered the bottle in larger proportions, creating an odd visual disconnect. The German woman was also incorrectly positioned in the scene. Grok clearly needs to perfect its understanding of proportions and spatial relationships, though they're moving in the right direction with these capabilities.
Imagine you're a film director wanting to visualize how Matt Damon might fit into your new production. Grok can simulate portrait manipulations:
 Original image |  “make him have long blonde hair” |
 “make him bald, wearing green sunglasses” |  “make the subject in the picture express immense joy” |
Overall, not bad! The facial modifications maintain a good level of accuracy while applying the requested changes.
Another promising application of Grok is creating composite images with fantasy elements layered over existing photos. This feature would be incredibly useful for creating eye-catching YouTube thumbnails:
 “add characters from the latest 2025 blockbuster movie [insert popular movie title] interacting with the landmark” |
 “transform this scene into a fantasy landscape with dragons flying in the sky and a magical castle in the distance” |
This is by far the most impressive update from xAI in terms of image editing capabilities, though at first glance it still falls slightly behind Google's recent release. xAI has shown quick development cycles, so I believe they will close this gap soon.
Grok 3's image editing capabilities are now available in the free version for all X users. To try it yourself, simply head to Grok, attach an image, make sure the 'Edit image' toggle is switched on, and enter your creative prompt.
INSPIRED BY AI
🎨 Creative of the Week
This week's pick may not have been made with AI (or at least that’s my guess), but it does promote an AI product—Perplexity. And it does so brilliantly, featuring Squid Game’s Lee Jung-jae.
No words needed—just hit play. You’ll relate instantly. 😀
Catch you next week!
— Radu
