OpenAI has introduced ChatGPT Images 2.5, a new image generation model. The update focuses on image quality, editing, speed, and creative control. OpenAI announced the update on September 8, 2026. The company says users create more than three billion images each week. This activity includes ChatGPT Images and GPT-Image models in the API. The new model delivers sharper details and richer textures. It also produces more natural lighting in generated images. Reference photos are now preserved more accurately during image creation. OpenAI also says the model follows editing instructions more reliably. This improvement applies across multiple editing turns. Generation latency has also dropped by up to 50% compared with Images 2.0.
More Precise Image Editing
One of the main changes involves image editing. ChatGPT Images 2.5 can make focused changes without altering unrelated details. Users can modify specific parts of an image while keeping the composition intact. This can help with products, backgrounds, text, and other visual elements. The model also improves consistency during longer editing sessions. Earlier changes are more likely to remain intact. Each new edit can build on the previous version without reducing image quality. This feature could benefit users working on detailed creative projects. It can also support marketing teams and businesses creating repeated visual assets.
ChatGPT Adds New Sketch Feature
OpenAI is also introducing a new feature called Sketch. It lets users draw directly inside ChatGPT. Users can create a rough drawing and use it as a visual reference. ChatGPT can then turn that sketch into a more complete image. The feature can help users communicate ideas without creating detailed artwork first. Users can sketch layouts, clothing concepts, room designs, or simple visual ideas. OpenAI says users can access the feature by typing “@Sketch” in ChatGPT.
Templates Make Image Creation Easier
The update also introduces templates for popular creative formats. Users can start with options such as posters and merchandise designs. Templates can reduce the need to begin with a blank canvas. Users can then add information, design elements, and preferred visual styles. ChatGPT Images 2.5 also improves its understanding of complex visual instructions. The model can handle detailed layouts and transparent backgrounds more effectively.
New API Models for Developers
OpenAI has introduced two new models through its API. They are called GPT-Image-2.5 Flare and GPT-Image-2.5 Sunburst. Flare focuses on quality, editing, and speed. OpenAI says it delivers higher-quality images than GPT-Image-2 with 50% lower latency. Sunburst targets premium creative workflows. It provides tighter control for detailed editing and production-ready visual work. These models can support creative, marketing, retail, media, and other business applications.
Availability and Safety
ChatGPT Images 2.5 is rolling out to ChatGPT, ChatGPT Work, and Codex users. OpenAI says it is available across desktop, mobile, and web platforms. The company is also continuing its image safety measures. These include prompt and image checks designed to reduce harmful outputs. OpenAI says it uses C2PA metadata and invisible watermarking. These measures can help identify images created with its tools. Overall, the update focuses on faster generation and stronger editing control. It also adds new creative tools for both everyday users and developers.
Explore our latest OpenAI news, GPT updates, and artificial intelligence developments for more updates.