GPT Image 1.5 I2I

GPT Image 1.5 image to image is OpenAI's visual edit mode, designed to take an existing image and modify it through natural language instructions. It preserves core structure, layout, and subject identity while applying controlled changes such as style shifts, object edits, lighting adjustments, or scene alterations. The model combines strong visual understanding with precise language conditioning, which allows it to make targeted edits without degrading overall coherence. GPT Image 1.5 I2I is well suited for iterative design workflows, creative refinements, and visual problem solving where accuracy and consistency are critical. Like the text to image variant, it is roughly 20 percent cheaper than GPT Image 1 on both input and output image costs, which makes high volume edit pipelines noticeably more affordable.

December 16, 2025

OpenAI

Image to Image (I2I)