ChatGPT Sketch turns your bad drawings into detailed AI images
The Verge Jay Peters ● Covered by 4 sources
OpenAI added Sketch to ChatGPT, so you can draw a doodle and ask it to turn that into an image. It can even fix a terrible mouse drawing into something realistic.
Based on reporting by The Verge, Jay Peters — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
OpenAI has added a new way to make images in ChatGPT: start with a sketch. The feature is called Sketch, and it sits inside ChatGPT as part of ChatGPT Images 2.5, which OpenAI announced on Tuesday.
The basic idea is simple. Draw a rough shape in a pop-up window, then tell ChatGPT what you want it to become. To open it, users type @Sketch into the chat box, and the drawing tool appears right there in the conversation.
The Verge says a quick test turned a bad cat doodle, drawn with a computer mouse, into a realistic photo after giving the model instructions. That is the appeal here. Not better prompting. A more forgiving prompt.
OpenAI is also adding another small but useful touch: people can now leave comments directly on parts of images. That makes the whole thing feel less like a one-shot generator and more like an editing workflow, even if the feature is still coming in through the side door.
My take — AI-written commentary, not fact-checked reporting
Sketch is exactly the kind of feature that makes AI image tools feel less like slot machines and more like software. Drawing badly and getting something useful back is a stronger pitch than another grand promise about creativity. OpenAI knows the real product is not the model; it’s the little buttons that make people actually use it.
Read more about this at: The Verge