OpenAI announced on Tuesday (16) the release of GPT Image 1.5, the latest version of its image-generation system within ChatGPT. The company says the new model delivers higher-quality results while operating up to four times faster than its predecessor, positioning it as a stronger competitor in the rapidly evolving AI image space.
With the update, users who select the “Images” tab in ChatGPT are presented with a redesigned creation interface. Instead of relying solely on written prompts, users can now choose from a range of built-in visual filters to generate images more quickly and intuitively.
The available preset styles include sidewalk mosaic, string art, retro TV, fair poster, outline, scribble, baroque, collectables, and pop art. These filters enable users to create stylised images without manually specifying each visual detail.
The new interface also features a “Discover something new” section, offering prewritten prompts that can be customized. Examples include celebrating a football victory, creating a holiday card, reimagining classic artwork, designing an album cover, or visualising oneself as a K-pop star.
Improved Editing Precision and Speed
OpenAI claims that GPT Image 1.5 significantly enhances image editing accuracy. The model is designed to modify only the elements requested by the user while preserving the rest of the image, allowing for cleaner and more controlled edits.
According to OpenAI CEO Sam Altman, the goal is for ChatGPT to function as a compact creative workspace. In a statement, the company explained that the update enables more realistic photo edits, better clothing and hairstyle simulations, and stylistic transformations that retain the original image’s character.
The model also enhances tasks such as adding, removing, merging, and rearranging visual elements. OpenAI claims GPT Image 1.5 follows instructions more reliably when both generating images from scratch and editing existing ones, reducing unexpected results.
Higher Visual Quality and Clearer Text
GPT Image 1.5 also brings visual improvements, including more natural rendering of multiple small faces and a noticeable reduction in the yellowish tint that users frequently criticized in earlier versions. OpenAI says overall color balance and realism have been refined.
Text rendering has also been improved. The model now handles smaller fonts and denser text more effectively, resulting in clearer and more readable generated images with written content. This aligns with broader industry efforts to improve AI-generated typography across different languages.
Availability
GPT Image 1.5 began rolling out to all ChatGPT users on Tuesday and is accessible through the “Images” tab. OpenAI states that the new features are supported across all models available on the platform, expanding advanced image-generation tools to its full user base.
 the release of GPT Image 1.5, the latest version of its image-generation system within ChatGPT.){kind=link}