G-Wagon AI Photo Trend: Copy-Paste Prompt for the Gas Station Night Shot

What the G-Wagon AI photo trend is

It is a photo of you with a black Mercedes G-Class, made by an image app from your own photo and a prompt. Through summer 2026 prompt sites published G-Wagon prompts for Gemini and ChatGPT, such as Auraprompt's car prompt list from June 29, 2026, where a man sits on the hood of a black G-Class on a foggy morning, and Media.io's collection of hood, garage and city night shots. This page covers one version: a gas station at night, the setting of our Photo with a G-Wagon template, with you leaning on the back of the car so the square body, the round spare wheel and the red tail lights show.

The copy-paste prompt

Open an image app that takes a photo with the prompt. In the Gemini app you upload the photo and ask for an edit in the same message, as Google's help page describes. Attach one full-length photo of yourself, then paste this:

Use my attached photo as the person. Keep my face, hair, skin tone and build exactly as they are. Late at night I lean back against the rear of a black Mercedes G-Class at an empty gas station, arms crossed. Dark bomber jacket, slim dark jeans, white sneakers. Light fog, red tail lights and canopy lamps glowing through the haze, wet reflective asphalt, deep shadows. Realistic photo, vertical 3:4, full length. No logos, no readable signs, blank license plate.

• The first two sentences come first on purpose. Image apps can drift toward a generic face unless they are told what to keep.
• "The rear" gets you the spare wheel and the tail lights, the two things that make the car read as a G-Class at night.
• "Light fog" rather than thick fog: the red glow needs some haze to show, but too much hides the car.
• "No readable signs" is there to keep fuel brands and garbled letters off the canopy. In our run the signs came out as panels with no readable names; without the phrase, a gas station scene can come with a real fuel brand's logo.

The run kept his face, the outfit and the pose. The station signs came out as panels with no readable brand names, though the pumps got number tags and label shapes; we blurred those too. Two things came out differently from the trend photos: the scene is brighter, with the canopy lamps lighting everything evenly, and the car is shot straight from behind rather than at an angle. For a darker shot, add "very dark scene, most light from the tail lights"; for the angle, add "seen from the rear three-quarter".

Variants: rain and neon

Both variants keep the same first two sentences and the same last line, so your face and the no-logo rule stay in place. Only the scene changes. We have not run the variants ourselves; only the main prompt above is tested.

Use my attached photo as the person. Keep my face, hair, skin tone and build exactly as they are. At night in steady rain I lean back against the rear of a black Mercedes G-Class at an empty gas station, hands in my jacket pockets. Wet black jacket, dark jeans, white sneakers. Rain streaks in the canopy lights, red tail lights reflected in puddles. Realistic photo, vertical 3:4, full length. No logos, no readable signs, blank license plate.

Use my attached photo as the person. Keep my face, hair, skin tone and build exactly as they are. At night I lean against the side of a black Mercedes G-Class in an empty parking garage lit by purple and blue neon light strips, arms crossed. Dark bomber jacket, dark jeans, white sneakers. Neon reflections on the glossy paint and wet concrete. Realistic photo, vertical 3:4, full length. No logos, no readable signs, blank license plate.

Which photo to attach

• Full length, head to shoes. The arms-crossed lean against the car only works when the app sees your whole body.
• Daylight or even indoor light on your face. The scene is dark, and a dark input loses your features twice.
• Face the camera with your hair the way you want it. The prompt keeps your hair as it is, so it can only copy what it sees.
• One person in the frame, with your arms visible. The pose crosses them, so the app has to know your build.
• Any casual outfit is fine: the prompt replaces it.

The one-tap version: Photo with a G-Wagon

Photo with a G-Wagon makes the shot from one full-length photo, with nothing to type. Its scene is darker and moodier than our prompt's: deep shadows, fog, red tail lights glowing through it and wet asphalt, with the plate covered by a black bar. The outfit is a dark bomber, slim jeans and white sneakers; on a woman's photo it is adapted to a women's look in the same pose. The station has no brand on it: a plain canopy with no signs or logos. The examples below are the template's results as delivered, with small edits by us: we blurred the car's badges and the plate's country band, and in the woman's photo we also blurred a small pump label reflected in a side window and painted out a small glowing mark on the asphalt. A run costs 1 photo, and it works on the free plan: new accounts get 3 photos free, with a watermark.

If the result looks off

• The face changed: run it again with a photo where your face is larger and evenly lit, and keep the "keep my face" sentence first.
• The car looks like another SUV: add "square boxy body, round spare wheel on the rear door, vertical tail lights".
• A station logo or letters appeared: AI text is garbled, so crop it out, blur it or run again.
• Your arms merged or a hand looks wrong: run it again; crossed arms are one of the harder poses for image apps.

For the same night mood with a sports sedan, see the BMW M5 fog night prompt. The Rolls-Royce Selfie moves the shot into a back seat, and Marker Outline on Your Car works on a photo of your own car.