CreativeWednesday, September 9, 2026· 2 min read

ChatGPT Sketch Makes AI Image Creation as Simple as Doodling

Source: The Verge AI

TL;DR

OpenAI’s new ChatGPT Images 2.5 adds Sketch, a feature that turns rough doodles into detailed AI-generated images. By letting people draw their ideas directly in ChatGPT, the tool makes visual creation more intuitive and accessible for everyday users.

Key Takeaways

  • 1ChatGPT Images 2.5 introduces Sketch, a doodle-to-image creation tool.
  • 2Users can type @Sketch to draw directly inside ChatGPT and guide the final image with text prompts.
  • 3The feature can transform simple mouse-drawn sketches into polished visuals, such as realistic photos.
  • 4New image commenting tools also make it easier to give precise feedback on specific parts of an image.

A more natural way to create with AI

OpenAI has introduced ChatGPT Images 2.5, including a new feature called Sketch that lets users start an AI image by simply drawing a rough doodle. Instead of relying only on text prompts, people can now communicate visual ideas in a way that feels immediate and creative.

By typing @Sketch into ChatGPT, users can open a drawing window, sketch an idea, and then describe how they want the final image to look. Early testing showed the tool turning a basic cat doodle drawn with a computer mouse into a realistic AI-generated photo.

This is a meaningful step for creative accessibility: people who may struggle to describe an image in words can now guide the AI with shapes, layouts, and visual cues. It also helps artists, designers, educators, and casual users move from idea to image faster.

  • Lower barrier to entry: Rough drawings can become detailed visuals.
  • Better creative control: Sketches and prompts work together to guide results.
  • More precise editing: Users can leave comments directly on parts of images.

Get AI Wins in Your Inbox

The best positive AI stories delivered to your inbox. No spam, unsubscribe anytime.