ChatGPT Had an Image Create Tool Update Earlier Today

ChatGPT had an image create tool update earlier today. The service called DALL-E is a playful blend of two influences which was the root for how the name came about.

Salvador Dalí is the famous surrealist artist known for his imaginative and dream like imagery and WALL·E is the robot from the Pixar movie and represents the artificial intelligence and the futuristic capabilities of non human AI sources.

With that said DALL·E doesn’t technically stand for anything, but more like an acronym. Like the ChatGPT which creates surreal, imaginative images, just like Dalí would; yet powered by technology like the little WALL·E robot. The idea behind the name is to capture the spirit of art and artificial intelligence which is what the ChatGPT service does, turns words into images.

So what was improved?

Inpainting (Image Editing / Generative Fill)

You can now click on parts of an image and edit or regenerate specific sections—like replacing an object or modifying just the background—without recreating the whole image. This gives you much more control.

More Accurate Prompt Interpretation

The tool is now better at understanding complex and detailed prompts, including spatial relationships, styles, and lighting. If you say “a USB drive plugged into the left side of a silver laptop,” it’s more likely to reflect that exactly.

Higher-Quality Images

There’s a noticeable boost in the sharpness, realism, and overall visual fidelity of the images generated. Textures, reflections, and lighting effects are now more natural-looking.

Faster Generation Time

Image creation is now quicker, allowing near real-time rendering of visuals—even for detailed prompts or specific camera angles.

Improved Consistency in Object Rendering

The model is better at producing recognizable and consistent objects, like realistic human anatomy, symmetrical faces, or technical devices such as laptops and USB drives.

Better Handling of Text in Images

While still not perfect, the system is better at rendering legible and relevant text within images—like signs, labels, or product branding.

Enhanced Style Control

You can more reliably request specific artistic styles (e.g., pencil sketch, watercolor, photorealism), and the system will stick to that aesthetic more closely.