Prompt Work

Create an image in the "Surreal Fish Doodle Landmark Phot...

Generate here

Use this prompt now

Cost

10 cr

Create an image in the "Surreal Fish Doodle Landmark Photo Collage" style.

Aspect ratio: 9:16.

Scene: [SUBJECT] [SUBJECT_ACTION], using [PRODUCT_OR_PROP], at [LOCATION]. Include [BACKGROUND_ELEMENTS]. Wardrobe styling: [WARDROBE_STYLE].

Photo base: use a realistic bright travel photograph with clear natural daylight, postcard-like scenic depth, readable sky and water or architecture, and a believable tourist-view perspective. Keep the landmark recognizable.

Illustration layer: overlay one oversized flat 2D illustrated subject across the foreground and middle ground. The subject should look pasted or screen-printed on top of the photo, with opaque flat color fills, thick black hand-drawn outlines, simplified hands or face, exaggerated scale, playful anatomy, and bold graphic wardrobe colors.

Creature layer: add two to five fantastical folk-art fish or flying fish around the main subject. Give them mismatched fins, wing-like shapes, round eyes, black tiger-like body stripes, red coral fins, cobalt-blue sections, yellow bodies, and simple black outlines. Let some creatures float in the sky, skim the waterline, or pass behind the main subject.

Doodle layer: add [ACCENT_SYMBOL] plus loose black marker loops, ink starbursts, blue water drops, curved strings, short splash dashes, and motion ticks as an independent graphic layer. Use these marks to create playful movement and connect the illustrated characters to the travel photo.

Text: keep typography minimal. If text is requested, use "[MAIN_TEXT]" only as a tiny hand-lettered travel note and "[SECONDARY_TEXT]" only as a sparse doodle label. Do not make large poster headlines.

Derive only the visual language from the style. Change the landmark, subject pose, outfit, props, and creature arrangement for every new image. Do not include watermarks, usernames, creator IDs, signatures, QR codes, platform marks, app UI, copied reference characters, exact reference poses, brand logos, readable brand marks, full vector-only backgrounds, glossy 3D cartoon rendering, photorealistic animals, dense text, or dark cyberpunk mood.

--- VARIABLES ---

[ACCENT_SYMBOL] — black marker loop, ink starburst, blue water drop, splash dash, curved string, motion tick, or tiny fish trail
[BACKGROUND_ELEMENTS] — blue sky, water, landmark structure, distant city, boats, coastline, clouds, bridge cables, buildings, or tourist-view scenery
[LOCATION] — recognizable travel landmark, bridge, bay, harbor, canal, cliff road, mountain pass, plaza, or observation deck
[MAIN_TEXT] — optional tiny hand-lettered travel caption or none
[PRODUCT_OR_PROP] — painted hiking staff, fishing line, travel bag, map, snack, camera, kite string, ferry ticket, umbrella, or oversized pencil
[SECONDARY_TEXT] — optional small doodle label or none
[SUBJECT] — oversized illustrated traveler, guide, fisherman, skateboarder, courier, musician, or explorer
[SUBJECT_ACTION] — leaning over a landmark, floating through the scene, pulling a stripe pole, balancing on a fish, chasing a creature, or bending around architecture
[WARDROBE_STYLE] — bold flat-color outfit, patterned grid shirt, oversized cap, sunglasses, striped socks, graphic shoes, raincoat, vest, or travel uniform

--- NEGATIVE PROMPT ---

watermark, username, creator ID, signature, platform logo, QR code, app UI, copied reference artwork, exact reference character, exact reference pose, original reference landmark copy, original reference outfit copy, original reference prop, logo, readable brand mark, full vector scene, blurred landmark, dark cyberpunk city, horror mood, dystopian scene, glossy 3D cartoon, plush mascot, photorealistic fish, realistic CG animal, heavy typography, dense text, corporate travel ad, sterile stock photo, messy unreadable collage

Creative GPT-Image-2 generation and editing for prompt-led campaign ideas. · 10 cr / 1K · refs

Upload sample imageAppears only for reference-capable models

Loading credits

Model

GPT Image

Type

Image

Related Works