OpenAI’s updated image generator can accept a rough sketch from the chat window and expand it into a finished image.
OpenAI released ChatGPT Images 2.5 on Tuesday with a new Sketch tool. The tool lets a user draw a rough doodle inside ChatGPT and then produce an image from that sketch. OpenAI also added the ability to leave comments on specific parts of an image to request edits.
OpenAI says ChatGPT Images 2.5 creates images with more natural lighting and richer textures. The Sketch feature appears when a user types @Sketch into the chat box and displays a drawing window. OpenAI built the update for ChatGPT, ChatGPT Work, and Codex users across desktop, mobile, and web.
Sketch gives builders an interactive way to shape image output without leaving the ChatGPT interface. Region comments allow operators to request changes to a specific part of an image. The latency reduction means ChatGPT can produce revised images more quickly during multi-turn workflows.
OpenAI made ChatGPT Images 2.5 available to ChatGPT, ChatGPT Work, and Codex users across desktop, mobile, and web. The version handles editing instructions across multiple turns, so teams can combine Sketch and region comments in the same workflow. Watch how those tools behave when a user sends a long series of image changes.
What matters
- OpenAI released ChatGPT Images 2.5 with a new Sketch feature that accepts doodles.
- Developers can combine Sketch with region comments to refine generated outputs in ChatGPT.
- Watch how the multi-turn editing improvements handle follow-up requests on generated images.
Why it matters
Watch how the multi-turn editing improvements handle follow-up requests on generated images.
This GenAI News article was prepared in original wording using reporting and materials published by The Verge. Source reference: https://www.theverge.com/ai-artificial-intelligence/991727/openai-chatgpt-images-2-5-sketch.
Drafted by the GenAI News review pipeline.