Portraits
Y2K Mall Snapshot
GPT Image 2 · @aronhouyu
Find a GPT Image 2.5 prompt for portraits, posters, product images, and more. Copy it and make it your own.
24 of 24 prompts
Reference examples from GPT Image 2 and Nano Banana. Each card credits its original model and creator.
Portraits
GPT Image 2 · @aronhouyu
Collages
GPT Image 2 · @Mind_Boticni
UI & social
GPT Image 2 · @Noor_ul_ain43
Collages
GPT Image 2 · @Sairah_0
UI & social
GPT Image 2 · @you1873118
Collages
GPT Image 2 · @Noor_ul_ain43
Collages
GPT Image 2 · @Mind_Boticni
Collages
GPT Image 2 · @Mind_Boticni
Illustrations
GPT Image 2 · @Naiknelofar788
Portraits
GPT Image 2 · @BubbleBrain
Collages
GPT Image 2 · @Shinning1010
Posters & covers
GPT Image 2 · @Mind_Boticni
Collages
GPT Image 2 · @Shinning1010
Collages
GPT Image 2 · @Taaruk_
Products & ads
GPT Image 2 · @you1873118
Illustrations
GPT Image 2 · @zhongying14
Portraits
GPT Image 2 · @GptImage2
Portraits
Nano Banana 2 · @GptImage2
Posters & covers
Nano Banana 2 · @GptImage2
Illustrations
GPT Image 2 · @GptImage2
UI & social
GPT Image 2 · @GptImage2
Portraits
GPT Image 2 · @GptImage2
Posters & covers
Nano Banana 2 · @GptImage2
UI & social
GPT Image 2 · @GptImage2
Start with the image you need: a product photo, a magazine cover, or a character illustration. Then describe the visible details that matter to you. Use the template as a starting point and replace the brackets before generating.
Name the subject, setting, and viewpoint. For a product photo, specify the material, background, and where the product sits in the frame. For a portrait, describe the pose, crop, and light source.
Choose a visual style, color palette, and lighting direction. If the image needs text, put the exact words in quotes and say where they belong. Check the spelling in the result.
When editing a reference image, name the part to change and the details to keep, such as the face, product shape, or composition. Compare the result before asking for another edit.
Create a [type of image] for [purpose]. Show [subject and visible details] in [setting]. Frame it as [viewpoint and composition], with [lighting] and [colors or materials]. Style: [photograph, illustration, collage, or another style]. Text: Place "[exact wording]" at [position], or use no text. For an edit: Change [one detail]. Keep [specific details] unchanged. Avoid: [unwanted elements].
Select an image to see the full prompt, or switch to the Prompts view to read the collection as text. Choose “Copy prompt” to use it in your own workflow, or “Use prompt” to open the image editor with the text filled in. Replace the subject, colors, and layout with your own details before generating.
No. The cards preserve the original model labels, including GPT Image 2 and Nano Banana. They are visual references and prompt ideas to adapt for GPT Image 2.5, not a set of verified GPT Image 2.5 outputs. Changing models or reference images can change the result.
You can start a new image from a text description. Upload a reference when your prompt asks to edit an existing image or retain a particular face, product, or layout. “Use prompt” transfers the text only; it does not attach the example image.