ChatGPT Images 2.5 arrives with faster generation, smarter edits
What's the story
OpenAI has launched an upgraded version of its image generation technology, ChatGPT Images 2.5. The new model comes with sharper details, faster generation times, and more precise editing capabilities than its predecessor. "Images 2.5 produces more natural lighting and richer textures," the company said in a statement. It is also better at preserving subjects in reference photos and following editing instructions across multiple turns.
Performance boost
'Sketch' feature introduced
OpenAI has also managed to cut down image generation latency by as much as 50% compared to Images 2.0.
This means users can generate images and refine their concepts much faster than before.
The company has also introduced a new "Sketch" feature that lets users draw directly in ChatGPT, which can then be used as a reference for image generation.
OpenAI is also introducing templates for popular image formats and a prompt sharing feature.
Developer access
Two versions of Images 2.5 released for developers
OpenAI has also released two versions of Images 2.5, namely GPT-Image-2.5 Flare and GPT-Image-2.5 Sunburst, for developers to build into their apps via OpenAI's API.
The former offers the same quality, editing, and speed improvements as the original model but at a lower latency.
It is ideal for various applications such as creator and social content, product experiences, visual search, rapid image prototyping, and high-volume generation tasks.
Premium tool
Sunburst version tailored for premium visual workflows
The second version, GPT-Image-2.5 Sunburst, is designed for premium visual workflows that require tighter control across edits.
It can be used for creative and editing workflows such as production-ready campaign creative or polished product imagery.
This gives developers more flexibility in using OpenAI's advanced image generation technology in their applications.