OpenAI Releases GPT Image 1.5 with Editing Improvements

OpenAI unveiled its latest image generation model, GPT Image 1.5, on Tuesday, making it available to all ChatGPT users worldwide and through its API. The update promises up to four times faster rendering speeds, enhanced precision in editing existing images, and improved adherence to user instructions, positioning the tool as a direct response to recent advancements by competitors like Google.

The new model builds on the previous GPT Image 1, released in April, by addressing key limitations in iterative editing. Users can now make targeted changes, such as adjusting lighting, facial expressions, or color tones, without the system overhauling unrelated parts of the image. This preserves elements like composition, facial likeness, and overall consistency across multiple modifications. OpenAI officials noted that the model excels at generating images from text prompts or photos, supporting tasks like adding, removing, or blending elements while maintaining spatial relationships and details.

A dedicated workspace enhances the user experience. Accessible via a new image tab in the ChatGPT sidebar on web and mobile platforms, it functions as a creative studio. Users can apply preset...