Published On 4/22/2026
|
Last update: 15:28 (Mecca time)
Over the past few hours, the American company for artificial intelligence technologies, OpenAI, has launched the latest version of the model for generating images from text under the name “ChatGPT Images 2.0”, making it a successor to the first version that was launched several years ago, according to a report by the “Digital Trends” technical website.
The new version of the model focuses primarily on accuracy and logical thinking, which makes it easier for it to generate text in images and maintain the style and accuracy of the images without significant hallucinations.
The new version of GPT Images Chat can search the Internet for sources and images similar to the images you wish to generate, in addition to its ability to generate more than one image by using only one command, according to what was stated in a technical report by “The Verge” website.
The model has also become able to generate images with a resolution exceeding 2048 pixels while maintaining the stability of the image contents and their consistency with each other, which is something that the previous version of the model faced difficulties with, according to a report by “The Verge.”
Understand the command instead of blind execution
The Digital Trends report confirms that the new version of the image generation tool now focuses on understanding the command directed to it first and then starting to implement it before presenting the final result.
During this time, it searches online for requests similar to what the user submits so that it can better understand the request and then provide the final result, unlike previous versions that were quick to implement requests and produce final images without paying attention to the instructions addressed to them.
This new logical thinking mechanism ensures the quality of the images generated by the tool, making it now closer to a digital creativity tool than just a tool that generates images randomly, according to the report.
The tool’s ability to preserve the general context of images and the method used to generate them makes it ideal for generating more than one image together and working on picture books, for example, as it can remember the details of images that it has generated previously.
Greater ability to use texts
The distinction of the new version of GPT Images Chat is evident in its ability to generate texts and integrate them directly into images in a very accurate manner while maintaining the quality of the image and text used in it.
This makes it ideal for everyday work that involves generating information images and graphics, such as restaurant menus, storyboard pages, or even architectural and home design drawings.
The report of the American technical website TechCrunch indicates the difficulty of distinguishing between images generated by the new GPT Images chat tool and images designed by human hands, which was not the case in the past, as it was possible to differentiate between images of artificial intelligence and images of humans.
Outperform competitors
A set of tests conducted by the “Arena.ai” platform, which specializes in evaluating artificial intelligence models, revealed a clear superiority of the “GPT Images 2.0 Chat” tool compared to the previous generation of the tool and even “Nano Banana 2”, which comes from Google, with a difference of more than 100 points in many aspects of the test.
The tests included measuring the tool’s ability at several levels, including the ability to modify images, whether one or more, as well as the ability to use texts within images and adhere to product names and commercial use, as well as the ability to generate realistic, cinematic images, and an imaginative image drawn with various artistic tools.
More uses in daily life
The new updates that the GPT Images Chat tool has received ensure its ability to benefit the end user in his daily life. Instead of spending a long time trying to direct commands and generate the correct images using previous versions, it has become much faster and more accurate with the new version of the tool.
The tool allows the user to generate 3 images for free every 24 hours, and he must subscribe to the ChatGPT Plus service if he wants to use it more than that, noting that accessing it does not require the use of any external applications or websites, and this can be done directly through the chat window in “ChatGPT Plus”, which enhances its usefulness for the end user who may need it to generate images in his various works.
