- Step1: Open GPT Image 2.5 and choose text-to-image or image-to-image.
- Step2: Describe the scene you want, or upload a JPEG, PNG, or WEBP photo up to 24MB.
- Step3: State the requested result, including what should change and which subject or framing should remain.
- Step4: Choose quality, aspect ratio, and output count, then generate the images.
- Step5: Review composition, subject details, text, and lighting before selecting a preferred version.
- Step6: Use the selected image as a reference and request one focused refinement at a time.