This Diorama Aquarium GPT Image 2.5 Prompt makes a detailed miniature world with a clear cutaway, tiny people, and controlled underwater light. Use it for a collectible-style concept image, a museum display idea, or an editorial illustration that needs scale.

Diorama Aquarium GPT Image 2.5 Prompt
Walnut Reef Case
Museum-quality cutaway diorama of a coral reef aquarium inside a walnut display case, tiny divers repairing a miniature glass tunnel, sea turtles, branching coral, soft shafts of daylight through rippling water, accurate scale details, 3:2 editorial product photograph, no text, no watermark
Settings: OpenAI GPT Image 2.5 Sunburst, 2K, 3:2, high quality, opaque JPEG, 1 sample.
Variations
Suitcase Sea Lab

Whimsical miniature aquarium diorama built inside a vintage suitcase, tiny marine biologist at a folding desk, glowing jellyfish, kelp forest, round porthole window, warm studio light, macro 100mm lens, 3:2 composition, no text, no watermark
Settings: OpenAI GPT Image 2.5 Sunburst, 2K, 3:2, high quality, opaque JPEG, 1 sample.
Underwater Lobby

Precise architectural diorama of an underwater public aquarium lobby, miniature visitors beneath a giant acrylic wave ceiling, patterned tile floor, bioluminescent fish shadows, long exposure look, clean gallery lighting, 3:2 composition, no text, no watermark
Settings: OpenAI GPT Image 2.5 Sunburst, 2K, 3:2, high quality, opaque JPEG, 1 sample.
Why this Diorama Aquarium GPT Image 2.5 Prompt works
- The cutaway instruction tells the model to reveal layers rather than make a normal aquarium photo.
- A walnut display case gives the miniature a physical boundary, which makes the scale easier to understand at once.
- Tiny divers and a glass tunnel create a human reference point. Their job also supplies a believable action.
- Coral, turtles, and rippling shafts of light give the water separate foreground, middle, and background details.
- The editorial product-photograph direction keeps the object centered and readable instead of turning it into a wide seascape.
Miniature scenes work best when the prompt gives the viewer a clear container. Without one, an aquarium diorama can drift into a life-size underwater scene. The walnut case anchors the result. It also adds a tactile material that contrasts with the glass and water.
The main prompt has a hard visual task: tiny people working inside a detailed reef. It names the action, the tunnel, and the creatures so the image has a hierarchy. The divers sit near the tunnel, while the coral and turtles fill the larger environment. That structure keeps the small figures from becoming an afterthought.
Soft daylight through moving water supplies atmosphere without hiding the craftsmanship. The requested 3:2 frame gives the case enough horizontal room for the reef while retaining an editorial crop. Sunburst at 2K is a sensible choice when small props, glass edges, and repeated coral forms need close inspection.
To make a fresh version, preserve the container and switch the ecosystem. A desert greenhouse in a suitcase, a lunar observatory in a cabinet, or a forest railway in a teacup all follow the same useful pattern. The object tells the eye where the miniature begins and ends. Then add one small figure with a task. A person repairing, cataloging, or carrying something makes the scale feel intentional. It also gives the scene a quiet story without needing a caption.
Do not overload a miniature prompt with unrelated props. Start with the container, the main environment, one human action, and the light. Add animals or architectural details only when they support that central picture. Macro lens language is especially useful for the suitcase variation because it explains the close point of view. The main image needs a more formal product-photograph instruction, so the display case stays legible on a grid card. If repeated objects come out too loose, name fewer types of coral and make the requested action more prominent. That gives the model more room for the important details.
More GPT Image 2.5 prompts
See more outputs in GPT Image 2.5 Prompts. Related examples include Trail Map Poster GPT Image 2.5 Prompt, Saffron Tea Label GPT Image 2.5 Prompt, and Citrus Serum Still Life GPT Image 2.5 Prompt. For a previous-generation comparison, visit GPT Image 2 Prompts.
For the maker’s own notes on this model, see OpenAI's image prompting guide.
Run this prompt yourself
Every example above was generated on Wiro. Paste the prompt, pick the model, and get your own result.