OpenAI rolls out ChatGPT Images 2.5, adding a sketch-to-image feature and cutting generation latency.
The AMW Read
Incremental image-generation feature update (sketch input, inline edit comments, latency cut) to an existing OpenAI case-study product, not a new entrant or structural shift.
OpenAI rolls out ChatGPT Images 2.5, adding a sketch-to-image feature and cutting generation latency.
OpenAI released ChatGPT Images 2.5 on September 9, 2026, alongside a new Sketch tool activated by typing @Sketch in the chat box. Sketch lets users draw a rough doodle directly inside ChatGPT and pair it with a text instruction to generate a finished image; users can also leave comments on specific parts of a generated image to request targeted edits, such as changing eye color. OpenAI says the update produces images with more natural lighting and richer textures, follows editing instructions more consistently across multiple turns, and cuts generation latency by up to 50% compared with Images 2.0. The update is live now for ChatGPT, ChatGPT Work, and Codex users across desktop, mobile, and web.
The release keeps image generation inside OpenAI's primary chat surface rather than spinning it out as a separate product, which matters because iterative visual editing — sketch input, targeted comments, multi-turn refinement — is exactly the workflow that has kept users bouncing to dedicated tools like Midjourney or Adobe Firefly. By narrowing the gap between a rough idea and a finished asset without leaving the chat window, OpenAI is treating image generation as a retention feature for its dominant consumer product rather than a standalone bet, and the latency cut addresses one of the more common complaints about slow image turnaround inside ChatGPT.
For builders on the Images API or Codex, faster generation and multi-turn edit consistency lower the cost of building image-heavy workflows on top of OpenAI's stack instead of a specialized provider. Standalone image-generation tools now face pressure to match not just output quality but the interaction model — sketch input and inline commenting — or risk losing casual users who default to whatever's already inside their chat app.
