OpenAI has made its image generation offering more accurate and consistent in its latest update to ChatGPT Images, as more enterprises and brands use AI image generation to help with design visualizations.
Updates will be released to all ChatGPT users and api As of GPT Image 1.5. The company said it is powered by GPT 5.2, which many early users found to be a powerful update for business use cases.
“Many people’s first experience with ChatGPT involves converting a text prompt into a picture,” said Fidzi Simo, OpenAI CEO of the application. In a Substack post“It’s a magical way to see what this technology can do, but the chat interface was not originally designed for this, Creating and editing images is a different kind of work and a place should be made for visuals,”
Business-friendly updates following precise editing and instructions
One of the biggest updates to ChatGPT Images is more targeted editing, even when the image originates on the chat platform rather than the API. Image generation models such as ChatGPAT Images, Google’s Nano Banana, and Stable Diffusion all involve prompt-based changes to AI-generated images, where the user can point to specific parts of the photo to change. But those features can sometimes be hit-and-miss.
With the update, OpenAI said the model better follows what the user wants “while keeping elements like lighting, composition, and the presence of people consistent across input, output, and subsequent edits.”
Users can instruct the model to perform most types of image editing, such as adding or subtracting an element, combining, blending, and transposing.
OpenAI said the model “follows instructions more reliably” than previous versions. It is also able to render text better and produce realistic, readable characters, whether dense or small. OpenAI updated the model to better produce smaller faces in photos of large groups of people.
According to OpenAI, “These changes work for both simple and more complex concepts, and the preset styles and views in the new ChatGPT Image feature are easy to use – no written prompts required.”
battle of image generators
OpenAI’s image model update follows Google’s The acclaimed Nano Banana Pro Image model, which received praise from the developer community.
The company must compete with other ever-growing, constantly improving image-production models aimed at attracting more enterprise users. And it’s not just Google that OpenAI has to contend with. Alibaba announced in August he quen-image Can produce readable text in both Chinese and English. Black Forest Labs Release flux.2Which also provides a robust, open-source image model.
<a href