OpenAI Launches ChatGPT Images 2.5 With Sharper Pictures And Faster Editing

 

 

OpenAI has launched ChatGPT Images 2.5, a new image generation model that promises sharper images, faster results, and more precise editing tools for users worldwide.

The Artificial Intelligence (AI) company says people already create more than three billion images every week using ChatGPT Images and its GPT-Image models via the API. With this update, OpenAI aims to make the creative process even easier and quicker.

The Images 2.5 feature is now rolling out to all ChatGPT users, whether they are paid users or not. On the other hand, GPT-Image-2.5 Flare and Sunburst are both available immediately through OpenAI's API.

 

Images 2.5 is now available to all ChatGPT, ChatGPT Work, and Codex users across desktop, mobile, and web platforms.

According to OpenAI, Images 2.5 produces more natural lighting and richer textures compared to earlier versions. It is also better at keeping the original subject looking accurate when working from reference photos, and it follows editing instructions more reliably, even across several rounds of changes.

The company said it has cut image generation time by up to 50 per cent compared to Images 2.0, allowing users to create and refine their ideas much faster.

 

Alongside the model update, OpenAI is introducing several new features in ChatGPT. One of these is called Sketch, which allows users to draw directly inside ChatGPT as a guide for the final image. This means someone can quickly sketch a room layout, an outfit idea, or even a simple doodle, and ChatGPT will turn it into a complete image based on that drawing.

The update also includes templates for popular formats such as flyers and product photos, making it easier for users to get started instead of working from a blank page. Users can now leave comments directly on images to request specific edits, and they can also share the prompts they used so others can try the same idea with their own photos and details.

 

 OpenAI said the new model is better at making focused edits, changing only the specific part a user asks for while keeping everything else the same, even in more complex images. This is useful for tasks like changing a single product, background, or piece of text without affecting the rest of the design.

The Images 2.5 model also keeps images more consistent across longer conversations. Earlier edits are less likely to change unexpectedly, and each new edit builds on previous ones without reducing the overall image quality.

In addition, Images 2.5 is said to better understand complex instructions, including detailed layouts and transparent backgrounds, while staying closer to a user's intended visual style.

 

For developers, OpenAI is releasing two new models through its API. GPT-Image 2.5 Flare offers the same improvements in quality and speed, and is designed for high-volume tasks like social content, product images, and quick prototypes.

Meanwhile, GPT-Image 2.5 Sunburst is designed for more detailed creative work, offering higher precision but with longer generation times. This model is intended for premium projects such as advertising campaigns and polished product photography.