OpenAI launches a tool for creating and editing images within gpt chat

Watch: Here’s How It Works – The Giphy Chat Photo Model

According to the OpenAI company, the new tool allows the creation of images from scratch, advanced editing of existing images and spot changes while preserving existing details. The launch announcement stated that the system makes more accurate edits, better understands the user’s intent, and maintains consistency between different products, a well-known problem that has received countless jokes in the artificial intelligence user community. Meanwhile, OpenAI also promises a significant improvement in speed, with image creation up to four times faster than previous releases.

Another innovation is the launch of a dedicated Images tab within the Gipity chat application and in the browser version. It is not only a work tool, but also an inspiration feed that presents ideas, styles and possible uses for creating images. The company explains that the goal is to make the process of creating images a more intuitive and enjoyable experience, which will encourage trial and error even among users who are not professional designers or artists.

From a technological point of view, OpenAI emphasizes that it is not a revolution but a deep polishing of the existing capabilities. The new model should be more reliable, adhere better to textual guidelines, and preserve small details even when performing a series of repeated edits to the same image. This ability is especially important for creators, publishers and designers who work with the same image over time and require high consistency.

However, initial field tests paint a more complex picture. Users who tried to make the same edits in both tools reported very similar results, with Google’s version sometimes being seen as visually more impressive. For example, editing a photo of a car at night and turning it into a day scene produced a reasonable result in both cases, but with a slight advantage for Google in the eyes of some testers.