OpenAI has made the Sketch tool available with the ChatGPT Images 2.5 model, which it introduced on September 8, allowing users to turn sketches into visuals. The company said that the generation wait time has been reduced by up to 50% compared with Images 2.0. In the tool selected by typing @ in the message box in the mobile app, users can provide sketch, subject, color, and style instructions together. The update brings edits that better preserve the distinctive features of people and objects in reference photos, natural lighting, rich textures, visual templates, and prompt sharing. When the background, product, or text is changed, other details are affected less; the accuracy and placement of successive edits and infographics are improved. Through a shared link, others can create a different version using their own photos and details.
Why it matters
This development gives weight not only to text prompts in image generation but also to the directing power of the sketch at hand; for this reason, it could affect how users who want to quickly visualize their ideas work. The reduction in wait time and improvements in successive edits lower the trial-and-error cost for those making detailed revisions to the same image. Better preservation of distinctive elements in reference images makes results more predictable in edits based on personal photos or specific objects. On the other hand, while the ability to share prompts and visual versions facilitates collaboration, it leaves open questions about how links will be managed and who will retain control of the original content in different versions. These points raise the question of how the tool will be assessed in collaborative production processes.
Background
ChatGPT is not a new name in the FikirPilot archive: we have published 17 news articles mentioning this name in the last 90 days; the most recent was dated September 9, 2026.