
GPT Image 2 is OpenAI's image generation and editing model released in April 2026, the generation that first paired its image output with reasoning. It improved instruction following, object placement, dense text and user-interface rendering, and it was announced alongside a thinking mode that can search the web and produce several distinct images from one brief. The API model, gpt-image-2, accepts text and image inputs, works through the images endpoints, supports inpainting and batch jobs, and produces images at flexible sizes. OpenAI now marks it deprecated in its model documentation in favour of the 2.5 variants, so it is a solid choice for existing integrations and pinned workflows rather than for new builds.
GPT Image 2
OpenAI画像
公開日2026年4月




