Original document
GPT Image 2 · en
Scene:
A square interior image held in patient stillness, with rain sliding down window glass, damp wood, muted blue-gray daylight, and one warm pool of lamplight or stove light inside. The room should feel inhabited but quiet, with time stretched by weather and silence rather than by spectacle. Show a wooden table pushed close to a rain-streaked window, with blue exterior light softening the edges of a ceramic cup and folded hands.
Subject:
Use one seated or standing figure, sometimes accompanied by a sleeping animal or simple domestic object, positioned so the window and the room share emotional weight. The human presence should be understated, contemplative, and integrated into the architecture. A solitary figure sits at the table in profile, not doing anything except watching rain collect and slide down the pane.
Important details:
The grammar depends on slow observation: wet glass, old timber, worn textiles, steam, puddled boots, enamel basins, and practical light. Preserve physical realism and spiritual quiet without turning the scene into fantasy, melodrama, or rustic cliche. Keep the room modest and tactile, with moisture and wood grain carrying the emotional tone.
Use case:
Long Gaze Through Rain Interior single study 1/9.
Constraints:
No readable text, no watermark, no logos, preserve exact 1:1 aspect ratio, no readable text, no obvious luxury styling, no dramatic storm destruction, no modern electronics dominating the room, no theatrical color grading, no broken anatomy, no low-detail background, no accidental text fragments.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.
