Original document
GPT Image 2 · en
A side-view screenshot shows two nearly identical, AI-generated street portraits against a light cyan background, displayed in two separate macOS-style image viewer windows. The layout comprises two vertical image panels, each within a rounded gray window with a top toolbar. The left window is titled "NanoBanana2," and the right window "ChatGPTImage2." Both panels present the same scene concept: a stylish young woman standing on a city street corner during prime time, framed from mid-thigh to head, her face obscured by soft rectangular blurs. She has long, wavy brown hair, a slender figure, and wears an off-the-shoulder, fitted beige long-sleeved top, high-waisted blue jeans, a delicate pendant necklace, and a large black leather bag with a gold chain slung over one shoulder. One hand holds a black smartphone near her waist. Both panels feature upscale, tree-lined city streets against a backdrop of warm sunlight filtering through the leaves, creating a shallow depth of field. The images include parked and moving vehicles, ancient stone buildings, and green traffic lights visible near the upper left corner of each picture. The two portraits are kept very similar but not identical: the left image has a slightly cooler, softer tone and a centered street perspective; the right image has a slightly warmer, brighter, and clearer tone, with stronger sunlight on the hair and a more dramatic blurring effect. Two viewer windows separated by a narrow vertical gap feature subtle shadows and a clean, technologically-driven contrast, capturing the visual contrast between two image-generating models.How to use
- Copy the complete original prompt, including any parameters.
- Replace the subject, text and composition with your own requirements.
- Use a compatible external tool; check its current access conditions and review the result.
Before you start
- Check whether a reference image is required. A compatible model label does not mean this example was generated with that model.
- Keep names, figures and visible text accurate; results can vary between model versions.
