- by Quentin Flambé
- on 16 Sep, 2025
Thirteen million people poured into one app in just four days—not to chat, but to remake their photos. That stampede hit Google’s Gemini app after the company rolled out Nano Banana, a text-to-image and image-editing feature that turns a simple prompt into glossy, shareable portraits. As usage spiked, executives called it a “full-on stampede,” and Google briefly put limits in place to keep the system from buckling.
The numbers are wild. The Gemini app jumped to more than 23 million users after the surge, and in that same burst, people ran 300 million extra images through the system, pushing it past the 500 million mark overall. The appeal is obvious: you type what you want, upload a selfie if you like, and the AI builds the shot around you—lighting, lens feel, background, wardrobe, even the tiny reflections in sunglasses.
At its core, Google Gemini with Nano Banana makes pro-style compositing feel like a phone trick. You don’t need Photoshop skills or studio lights. Tell it “golden hour, shallow depth of field, warm tones,” and it understands. Ask for “3:4, high angle, overhead,” and it nails the camera setup. The experience is less like using software and more like giving a director short instructions.
What Nano Banana does—and why everyone’s using it
Nano Banana handles two big jobs. First, it can generate new images from scratch using your text. Second, it can edit existing photos—swapping settings, adding props, or moving you into a totally different scene while trying to keep your face and expression intact. That one-two punch is why it’s catching fire on Instagram and TikTok. Pre-wedding shoots without a booking? Done. Polaroid throwbacks with grain, flash bloom, and a fake celebrity cameo? Also done. Even miniature-style portraits that look like collectible figures are trending.
Google says the crush of demand forced temporary caps during peak hours. It’s the classic viral-tool problem: the moment people see their friends post a slick stadium portrait or a moody streetwear cover, they want the same look. Under the hood, the system leans on text cues for lighting (“soft sunbeams,” “cinematic warm color grade”), lens and depth (“50mm prime, shallow DOF, creamy bokeh”), and scene control (“dim basement garage,” “ocean balcony, breezy”). The more specific your prompt, the cleaner the output.
Plenty of people stick to quick one-liners. Others write prompts like mini shot lists, tagging angle, aspect ratio, wardrobe, pose, and a short vibe sentence. Both work, but detailed prompts tend to produce more consistent results, especially when you want the AI to respect your outfit, hair, or tattoos.
The prompts men are using to land viral looks
Three styles are leading the trend right now, each built around a clear pose, mood, and setting. Here’s how creators are describing them—and how to pull them off yourself.
- Lamborghini garage pose
The look: A relaxed, overhead 3:4 full-body shot on the hood of a red SUV in a low-lit basement garage. The vibe is moneyed but understated—soft light streaks, warm grading, shallow DOF.
Wardrobe: Open-collar white shirt, brown trousers, polished black shoes, leather-strap watch. A forearm tattoo peeks out.
Key prompt cues: “Overhead high angle, 3:4 full body, dim garage, soft sunbeam reflections on red SUV, warm cinematic grade, shallow depth of field, creamy bokeh, ultra-detailed face and clothing, billionaire mood.”
Try a one-line starter: “Relaxed on the hood of a red performance SUV, overhead high angle, warm cinematic light, shallow DOF, crisp white shirt, brown trousers, leather watch, forearm tattoo visible.” - Romantic beachside balcony
The look: A clean portrait on a balcony with ocean waves behind you. You’re mid-action—adjusting sunglasses with one hand while holding a partner, breeze in the shirt, bright and airy color.
Wardrobe: Navy shirt, white trousers, loafers, sunglasses.
Key prompt cues: “Hyper-real portrait, beachside balcony, ocean in background, breezy and romantic mood, natural daylight, subtle rim light, skin tones natural, fabric movement.”
Starter: “Portrait on ocean-view balcony, adjusting sunglasses with one hand, holding partner, navy shirt, white trousers, natural daylight, bright and breezy, soft rim light, realistic skin texture.” - Football focus profile
The look: Tight profile, head angled up, a red-and-white match ball balanced on your forehead. Confident posture, jawline defined, deep blue backdrop for punchy contrast.
Wardrobe: Pro kit—fitted jersey and shorts.
Key prompt cues: “Profile view, ball balanced on forehead, focused gaze, elongated neck, gently closed lips, slightly raised chin, deep blue backdrop, dramatic light with soft shadow shaping.”
Starter: “Profile portrait, match ball balanced on forehead, looking upward with full concentration, pro jersey and shorts, deep blue background, dramatic soft shadows, sharp facial detail.”
Those are the viral staples. But the trend is bigger than three setups. Here are other styles popping up in feeds and group chats:
- Retro Polaroid with celebrity cameo: Add “instant film border, on-camera flash, color shift, slight blur” and a clean pose. For the cameo feel without impersonation issues, use “celebrity-style friend” or “famous-lookalike friend” rather than naming a public figure.
- 3D collectible figurine: “Tiny desk figurine, matte plastic texture, dramatic top light, studio backdrop, exaggerated features but realistic face identity.” It’s a playful way to make merch-like portraits.
- Streetwear magazine cover: “Low-angle street shot, neon spill, reflective puddles, bold type layout space, teal-and-orange grade, 35mm feel, motion blur.” Keep brand logos generic to avoid takedowns.
- Groom’s pre-wedding editorial: “Tailored suit, soft window light, reflective watch face, boutonnière detail, shallow DOF, hotel corridor.” A quick way to plan or preview a look before an actual shoot.
- Gym power pose: “Chalk dust in air, hard side light, metal textures, sweat detail, grainy black-and-white, mid-squat angle.” Simple, dramatic, and cheap to replicate in real life later.
If you’re new to prompting, think like a photographer. You’re telling the AI your lens, light, angle, color, and styling—in short sentences. The system fills in the rest.
Want more consistent results? Borrow this structure:
- Subject: your role and pose (“male model, relaxed pose holding sunglasses”).
- Camera: angle and framing (“overhead high angle, 3:4 full body” or “tight profile, 1:1”).
- Light: source and mood (“soft sunbeams, warm cinematic grade” or “dramatic single key light, soft shadows”).
- Setting: clear place cues (“dim underground garage” or “beachside balcony with ocean waves”).
- Wardrobe: exact pieces and color (“open-collar white shirt, brown trousers, loafers”).
- Detail: textures and finishing (“creamy bokeh, realistic skin texture, natural reflections”).
Pro tip from creators: add a short identity line when editing a real selfie, like “match subject’s face to the input photo, keep facial features accurate.” It nudges the model to preserve your look—eyes, hairline, beard shape—instead of drifting toward a stock face.
On aspect ratio, 3:4 flatters most full-body portraits. Go 4:5 for Instagram feed, 9:16 for Stories/Reels, and 1:1 for profile pics. If you want that magazine feel, mention a “35mm prime” or “85mm portrait” look; it guides perspective and background blur.
Lighting words do a lot of heavy lifting. “Golden hour,” “soft rim light,” “window light,” “hard side light”—they all give the model a strong baseline. If your shot feels flat, add a rim light or “specular highlights on metal and paint” for vehicles. If skin looks plastic, use “realistic skin texture, subtle pores, no plastic sheen.”
What about realism? The tool can push to “hyper-real” easily, but you don’t have to max it out. For a natural look, scale back with phrases like “true-to-life color,” “minimal retouch,” and “gentle contrast.” For a fashion edit, go heavier: “studio polish, high micro-contrast, glossy highlights.”
Creators also report better results when they upload a clean, front-lit selfie. Blurry or dim inputs make it harder to keep your face intact. Glasses, hats, and strong backlight can confuse outlines—say “preserve sunglasses” if you really want to keep them. And if the system keeps changing your shirt or tattoo, call it out: “lock wardrobe and forearm tattoo.”
Tools like this raise fair questions. Where’s the line between fun portrait and misuse? Google runs safety filters, so you’ll hit blocks on sexual content, violence, and obvious impersonation attempts. Models also struggle with exact branded items and team kits; references may look “inspired by” rather than licensed. That’s by design. If you’re posting publicly, it helps to label images as AI-made, skip real logos, and steer clear of public figure likenesses.
There’s also the watermark conversation. Google has promoted watermarking tech in parts of its AI stack, and the industry is moving that way. Don’t assume every image carries a perfect, permanent mark, though. If authenticity matters (weddings, ads), say it’s AI-generated up front and keep your originals.
Why is this trend winning now? Three reasons. First, the results look good on a phone—tight framing, big faces, punchy light. Second, the prompts are short. You don’t need a 20-line script to get something worth posting. Third, it fits into the way people already create: take a quick selfie, pick a vibe, share in minutes. The gap between idea and output is tiny.
It helps that the tool covers more than glossy portraits. People are turning out retro postcards, travel album fillers, and even moodboards for haircuts and beards. The “see it before you do it” use case—try a fade, test a suit color, preview a tattoo—is sneaky-powerful. It saves time and avoids buyer’s remorse.
Of course, pressure comes with popularity. Those temporary usage limits are a sign of scale, not a bug. Expect queues during big drops or weekend rush. Google says the team is tuning capacity to ride the waves, which is what you’d expect with any blockbuster feature.
How does this stack up against older tools? Midjourney still shines for raw artistry. Adobe’s tools are great if you live in a pro editing workflow. The twist here is the phone-first flow: one app, one prompt, one share button. For most people, that’s enough to win the day. If you’re a creator, it’s an easy way to upgrade thumbnails, posters, or podcast art without commissioning every single shot.
Want tighter control? Nudge the model with these small additions:
- “Natural color grading, subtle film grain” for a timeless look.
- “High dynamic range, soft highlights” to keep white shirts from blowing out.
- “Reflections on paint and glass” for cars and sunglasses.
- “Skin tone match to input photo” to avoid odd color shifts.
- “Consistent identity across variations” when generating multiple looks.
If you’re aiming for a big reveal—say, a birthday post or engagement teaser—generate three or four versions and pick the most convincing one. Look closely at hands, ears, and fine textures. If something feels off, run a small edit pass, or regenerate with the same seed to keep the lighting but fix the flaw.
One more thing about trends: they evolve fast. The garage pose will get copied to death. So will the balcony shot. Add a twist. Swap the car for a vintage motorcycle. Trade the ocean for a cliffside trail at “blue hour.” Keep the same prompt structure but move the light source or change the focal length. You’ll stay ahead of the copycats while riding the same wave.
The bottom line for anyone curious: the recipe is simple. A clear pose, a clean wardrobe callout, a specific setting, and two good lighting cues. Stick to that, and Nano Banana does the heavy lifting while keeping your face recognizable and your feed ready to go.