Studio Ghibli portraits were generated more than 50 million times within weeks of breaking through in early 2026. That number is worth sitting with for a second, because it explains everything that’s happened since. The trend cycle didn’t slow down after Ghibli. It got faster, weirder, and considerably more specific about what it wants from the AI.
Five trends are currently dominating Instagram, TikTok, and Pinterest. Each one runs on a single well-written prompt. Here’s what’s working right now, why it’s working, and the exact text to paste into ChatGPT, Gemini, or Nano Banana to get it.
1. The A24 Movie Poster
The trend: turn your portrait into a moody indie film poster, complete with fake critic quotes and a title treatment that looks like it belongs outside a Sundance screening.
Why it’s everywhere: the appeal isn’t really about the poster. It’s about the implied narrative. A single portrait becomes a whole imagined film — a backstory, a tone, a tagline — and that imaginative leap is what makes people stop scrolling. It also taps into a specific aesthetic vocabulary that A24 spent the last decade building into a recognizable brand, which means the AI has a huge, consistent visual reference to draw from.
The prompt:
“Create a dramatic indie movie poster using this portrait as the lead actor. Add cinematic shadows, subtle film grain, elegant title typography, made-up critic quotes in italic, and a moody color grade inspired by modern A24 films — leave room at the bottom for a fake tagline and release date.”
The variation worth trying if everyone in your feed has already done the A24 version: swap the reference point for Hong Kong cinema instead. Teal-and-magenta color grading, Chinese title typography with English subtitles, and the slow-shutter blur associated with Wong Kar-wai’s films produces something that reads as a theatrical one-sheet rather than a filter, and almost nobody else in your feed will have made one yet.
2. The Toyification / Action Figure Trend
The trend: turn yourself into a collectible toy, complete with blister-pack packaging, accessories, and toy-line branding.
Why it’s everywhere: everyone gets to be the main character of their own toy line, and the format is endlessly customizable through the accessories alone. A person who posts about coffee and dogs gets a figure that comes with a tiny latte and a tiny dog. A person who posts about their job gets a figure boxed under a toy-line name like “CAREER LEGENDS.” The personalization is the whole joke, and it’s different for every single person who tries it.
The prompt:
“Transform the subject into a stylized collectible toy with glossy plastic textures, exaggerated wide-set eyes, miniature accessories, and premium retail packaging. Add dramatic studio lighting and make it look like a limited-edition designer figure sitting on a velvet display pad. Name the toy line ‘[YOUR TOY LINE NAME]’ and include the accessories: [LIST 2-3 ITEMS THAT REPRESENT YOU]. Add ‘Limited Edition, Series 2’ text to the packaging.”
Note: The more specific the accessory list, the better the result. “Tiny laptop, coffee cup, and miniature dog” produces a sharper, funnier image than “some accessories.”
3. Disposable Camera Nostalgia
The trend: deliberately make a photo look worse — film grain, light leaks, soft blur, faded color — to make it feel like a memory instead of a post.
Why it’s everywhere: this is a direct reaction to several years of AI images that were too clean, too symmetrical, and too obviously generated. Gen Z creators who grew up surrounded by polished influencer photography are now actively choosing the opposite. The imperfection is the point. A slightly blurry, slightly overexposed photo with a light leak in one corner reads as authentic in a way that a technically perfect portrait doesn’t.
The prompt:
“Turn this photo into a faded late-90s disposable camera memory. Add dusty film grain, soft motion blur, uneven lighting, tiny scratches, and a warm sunset glow leaking from one corner. Make it feel accidental and nostalgic instead of polished.”
Note: The era you specify matters more than it seems like it should. “Late-90s” triggers a specific warm, slightly yellow color palette and a particular grain pattern that AI models have clearly been trained on extensively. Try swapping in a different decade — “early 2000s digital camera” produces a noticeably different, cooler-toned result with visible compression artifacts instead of film grain.
4. The Scrapbook / Diary Page
The trend: layer your photo with torn paper edges, handwritten notes, tape, coffee stains, and doodles so it looks like a physical diary page that got scanned.
Why it’s everywhere: it’s the toyification trend’s quieter cousin — instead of turning you into an object, it turns a single photo into a scene with implied history. The torn edges and tape suggest the photo was physically handled, kept, and revisited. For travel and relationship content specifically, that implied physicality does a lot of emotional work that a clean photo can’t.
The prompt:
“Turn this image into a handmade scrapbook page with torn notebook edges, handwritten notes taped at angles, coffee ring stains, faded magazine cutouts layered around the photo, small doodled flowers, and soft pastel textures underneath — make it look like a diary page that was scanned and uploaded.”
Note: This one rewards layering. If the first result looks too clean, ask for “more visible tape, at least one coffee ring overlapping the photo itself, and handwriting that looks rushed rather than neat.”
5. 90s Yearbook Photo
The trend: laser backgrounds, crimped hair, denim everything — the specific teal-and-purple gradient that only ever existed in school portrait studios.
Why it’s everywhere: like the disposable camera trend, this is nostalgia-driven, but it’s nostalgia for an aesthetic most people under 35 didn’t actually experience firsthand. The laser background in particular has become a kind of internet shorthand — instantly recognizable, slightly absurd, and impossible to take seriously, which is exactly why people love using it on themselves.
The prompt:
“Transform into an authentic 1995 high school yearbook photo. Classic 90s hairstyle, denim jacket, laser beam background in teal and purple gradient, soft studio lighting with slight vignette, and that unmistakable white border around the edge of the photo.”
The Pattern Underneath All Five
Every one of these trends works for the same reason: each prompt specifies what is in the photo, the style and mood, the background, the colors, and the lighting — in that order. The prompts that go viral aren’t the ones asking for something vague like “make this cooler.” They’re the ones that read like actual photography direction, because that’s effectively what they are.
The other thing all five trends share: every prompt above works across ChatGPT’s image generation, Google Gemini (including Nano Banana), Adobe Firefly, and FLUX with little to no modification. The platform matters less than the specificity of what you ask for. Try the same prompt across two different tools and compare — the differences in how each model interprets “moody” or “nostalgic” are often more interesting than the trend itself.
About the Author
Your 24-year-old cousin who has generated herself as a Ghibli character, an action figure, and a 1995 yearbook photo in the same week, and now refers to her camera roll as “the multiverse.”
