In a recent LinkedIn post, Federico Donatone explores the significant advancements introduced with OpenAI’s latest image generation model, ChatGPT Images 2.5. Donatone breaks down the key changes, emphasizing how these updates are poised to revolutionize creative workflows, particularly in advertising and visual content creation.
Revolutionary Speed and Precision in AI Image Generation
Federico Donatone begins by highlighting the dramatic increase in efficiency, noting that OpenAI claims the new model is up to 50% faster than its predecessor. This speed enhancement is crucial for businesses operating in fast-paced markets. He points out a fundamental shift in how AI image editing now functions:
“Until now, every time you asked AI to fix one detail, it redrew the whole picture.”
This previous limitation meant that even minor adjustments could lead to unintended alterations across the entire image, often requiring extensive rework. Donatone stresses that this is no longer the case with Images 2.5.
Contextual Edits and Persistent Features
A key innovation discussed by Donatone is the ability to perform highly localized edits. This means users can now target specific areas of an image for modification without affecting other elements. As Federico Donatone notes, this allows for more nuanced control:
“You can comment on one part of a picture and only that part changes.”
Furthermore, Donatone emphasizes the enhanced consistency offered by the new model. This is particularly important for elements like faces, pets, and products, which are critical in branding and marketing. He explains:
“Faces, pets and products stay the same across repeated edits.”
This persistence ensures that brand elements and key visual components remain stable throughout the editing process, streamlining the creation of consistent visual assets.
Introducing New Tools and Developer Options
Federico Donatone also introduces a novel tool that further enhances user control and creative input. The new ‘Sketch’ tool allows users to draw directly within ChatGPT to visually communicate their desired changes, bridging the gap between conceptual ideas and AI execution. This visual feedback mechanism is a significant step towards more intuitive AI interaction.
For developers, Donatone highlights the availability of two distinct versions of the model: Flare, optimized for speed, and Sunburst, designed for maximum precision. This tiered approach allows developers to select the best tool for their specific application needs, whether prioritizing rapid iteration or meticulous detail.
The Impact on Ad Visual Creation
Concluding his analysis, Federico Donatone draws a direct line from these technological advancements to practical business applications, particularly in advertising. He posits that the ability to make precise, localized edits quickly can drastically reduce production times. Donatone challenges businesses with a pointed question:
“So an ad visual that took days of back and forth with a designer takes one edit. Is your team still making ad visuals by hand?”
This rhetorical question underscores the potential for AI-powered image editing to disrupt traditional workflows, offering significant time and cost savings for marketing and creative teams.
📝 About This Content
This article is based on insights shared by Federico Donatone on LinkedIn.
📅 Originally posted on September 9, 2026 | View original post on LinkedIn →