GPT Image 2 launches as native multimodal image generator — lets users create and edit images directly without switching tools

GPT Image 2, a new image generation tool, launched at gptimg.co. It is built as a native multimodal system, meaning it combines text understanding and image creation in a single interface rather than requiring separate tools or API calls.

Unlike traditional image generators that only take text prompts, this tool can also accept images as input—letting users upload a photo and ask the AI to modify, extend, or reinterpret it. Users can work through iterations (describe, generate, refine) without leaving the interface.

For designers, marketers, and content creators, this eliminates the friction of copying prompts between ChatGPT and a separate image tool. Work stays in one place, reducing context-switching and speeding up creative workflows—especially for projects requiring multiple rounds of edits.