80s AI Photo Prompts: 8 Retro Looks Tested on 3 Models

8 80s AI photo prompts, from laser yearbook portraits to mall glamour shots, run on Nano Banana Pro, Nano Banana 2.1 and GPT Image 2.5. Which keeps your face?

Por BulkImagen Team10 min de lectura
Faded instant photos of laser backdrops and palm trees scattered on a pink 1980s table next to roller skates, a cassette tape and orange foam headphones

The 1980s photo has been one of this fall's big AI photo trends: upload a selfie, ask for a laser-background yearbook portrait or a soft-focus mall glamour shot, and post the result next to the original. Most prompt lists stop at the prompt. We wanted to know what actually comes back, and which model still gives you your own face under the big hair.

So we wrote eight 80s AI photo prompts, from the yearbook portrait to a Walkman street photo, and ran every one on three image models with the same selfie: Google's Nano Banana Pro and Nano Banana 2.1 (the image models behind Gemini) and OpenAI's GPT Image 2.5 (the model behind ChatGPT Images 2.5).

Below: the eight prompts, the side-by-side results, a close look at the faces, and how to make the same looks for a whole group in one go.

A grid of eight 1980s portraits of the same bearded man around his original selfie: laser yearbook photo, mall glamour shot, aerobics, prom, arcade Polaroid, roller rink, Walkman street photo and studio portrait
One selfie, eight looks, all from GPT Image 2.5

Genera por lotes

Un prompt por línea, cientos de imágenes, un ZIP para descargar.

Pruébalo gratis

How we tested#

  • One selfie, made for the test. The man in every picture is fictional: we generated his selfie with an AI model rather than use a real person's photo. It is a casual phone shot in a kitchen, with short curly hair and a short beard, the kind of photo people actually upload.
  • Eight looks, eight prompts. Each prompt describes the setting, the outfit, the lighting and the film or camera look, and ends with the same identity line: keep the person exactly as they are: same face, same features, same skin tone, same age, same hairstyle.
  • Three models, as available in October 2026: Nano Banana Pro and Nano Banana 2.1 (Google), and GPT Image 2.5 (OpenAI, the Flare variant).
  • Same settings everywhere: the selfie as the reference image, 2K resolution, 3:4 portrait, one attempt per prompt and no cherry-picking. All images rendered on the first try.

One honest note: our first versions of the prom and studio prompts described women's clothes (a satin prom dress, a bow blouse), and all three models dressed our male subject in them without hesitation. We rewrote those two prompts with clothes that fit the person and re-ran them on all three models. The lesson is in the tips below.

One run per prompt is a small sample, so read the results as tendencies. The main pattern repeated across most of the eight looks.

Results at a glance#

Model Keeps the face Follows the outfit Era look Watch out for 2K credits
GPT Image 2.5 (OpenAI) Closest. Kept the beard and his age in all 8 Very good, every detail Strong film color and texture Costs the most 10
Nano Banana 2.1 (Google) Close. Kept the stubble, one face drifted Good, sometimes skips a detail Most believable crowds and places Made-up lettering on signs 6
Nano Banana Pro (Google) Weakest. Shaved and younger in 6 of 8 Good Clean, polished Turns adults into teenagers 6

The short version:

  • To look like yourself: GPT Image 2.5, then Nano Banana 2.1, then Nano Banana Pro.
  • For the best value: Nano Banana 2.1 keeps the face nearly as well at 6 credits.
  • Nano Banana Pro made great 80s pictures, just not always of the same man.

This is close to the opposite of our Halloween photo test, where Nano Banana Pro kept a woman's face best and GPT Image 2.5 made her look older. Different selfies, different winners: test your own photo before running a batch.

The 8 80s AI photo prompts#

Each prompt below is exactly what the models received. Change the outfit details freely; keep the last two sentences.

1. Laser yearbook portrait#

Prompt
1980s yearbook portrait: turn the person from the reference photo into an authentic 1986 high school yearbook portrait. Head-and-shoulders pose turned slightly to one side, big teased hair styled in a typical 1980s way, pastel sweater with shoulder pads, posed in front of a classic purple and blue laser-beam studio backdrop. Soft studio flash, slightly faded color film print, fine grain. Keep the person exactly as they are: same face, same features, same skin tone, same age. No text, no logos.
The yearbook prompt on Nano Banana Pro, Nano Banana 2.1 and GPT Image 2.5: the man with big curly hair and a pastel sweater in front of a purple and blue laser backdrop
Laser yearbook portrait, left to right: Nano Banana Pro, Nano Banana 2.1, GPT Image 2.5

The signature look of the trend, and every model nailed the laser backdrop and the sweater. The difference is the face. Nano Banana Pro took "high school yearbook" literally: no beard, and he looks about seventeen. Nano Banana 2.1 kept the stubble but took a few years off. GPT Image 2.5 kept the beard, the age and the face, and gave him a perfect 80s mullet.

2. Mall glamour shot#

Prompt
1980s mall glamour shot: turn the person from the reference photo into a 1980s shopping-mall glamour portrait studio photo. Dramatic soft-focus haze, big voluminous hair, sequined jacket with large shoulder pads, oversized earrings, hand resting under the chin, chin tilted up, smoky pastel backdrop. Heavy diffusion filter, glowing highlights, 1980s color film print. Keep the person exactly as they are: same face, same features, same skin tone, same age. No text, no logos.
The mall glamour shot prompt on three models: the man in a sequined jacket with big hair and earrings, hand under his chin, against a hazy pastel backdrop
Mall glamour shot, left to right: Nano Banana Pro, Nano Banana 2.1, GPT Image 2.5

The most fun result of the test. All three did the hand-under-the-chin pose, the sequins and the haze. Nano Banana Pro shaved him again; the other two kept the beard. Nano Banana 2.1's version, with the starry backdrop, looks most like a real 1980s studio print.

3. Aerobics video#

Prompt
1980s aerobics: turn the person from the reference photo into a still from a 1984 home aerobics workout video. They wear a shiny high-cut leotard over tights, leg warmers and a sweatband, mid-move with arms raised in a bright studio with a pastel grid floor and mirrors. VHS tape look with soft color bleed, bright flat lighting. Keep the person exactly as they are: same face, same features, same skin tone, same age, same hairstyle color. No text, no logos.
The aerobics prompt on three models: the man in a shiny leotard, tights, leg warmers and a sweatband with arms raised in a mirrored studio
Aerobics video, left to right: Nano Banana Pro, Nano Banana 2.1, GPT Image 2.5

GPT Image 2.5 is the only one that looks like a frame from a tape, with soft color bleed and a noisy edge. Nano Banana 2.1 added a class of other people in the mirror, which makes it feel real.

4. Prom night#

Prompt
1980s prom photo: turn the person from the reference photo into a posed 1987 prom night photograph. They wear a powder-blue tuxedo with a ruffled tuxedo shirt, a matching bow tie and cummerbund and a flower boutonniere, standing in front of a shimmering silver tinsel backdrop with balloons. Direct flash, slightly washed-out color print from a drugstore photo lab. Keep the person exactly as they are: same face, same features, same skin tone, same age, same hairstyle. No text, no logos, no signs, no date stamp.
The prom prompt on three models: the man in a powder-blue tuxedo with a bow tie and boutonniere in front of silver tinsel and balloons
Prom night, left to right: Nano Banana Pro, Nano Banana 2.1, GPT Image 2.5

GPT Image 2.5 got every item on the list: ruffled shirt, bow tie, cummerbund, boutonniere. Nano Banana 2.1 gave him a plain tuxedo shirt instead of the ruffles. Nano Banana Pro got the outfit right but, once more, a younger, clean-shaven face.

We added "no signs, no date stamp" to this prompt after the first version came back with a "PROM '87" banner on one model and a printed date in the corner on another.

5. Arcade Polaroid#

Prompt
1980s arcade Polaroid: turn the person from the reference photo into a 1980s instant Polaroid snapshot taken in a neon-lit video arcade. They lean against an arcade cabinet in a denim jacket with pins, glowing screens and neon tubes behind them. Instant film color shift, soft focus, white Polaroid border, slight light leak. Keep the person exactly as they are: same face, same features, same skin tone, same age, same hairstyle. No text, no logos, no readable game titles.
The arcade Polaroid prompt on three models: instant photos with white borders of the man in a denim jacket with pins in a neon arcade
Arcade Polaroid, left to right: Nano Banana Pro, Nano Banana 2.1, GPT Image 2.5

The look where all three kept the face best: no big hair, no costume, just a jacket. All three drew the white instant-photo border. Nano Banana 2.1 painted a famous game's logo and character onto the cabinet next to him, despite "no readable game titles"; we blurred it in the comparison above. Check backgrounds before you post: a real brand in your picture is not something the prompt can promise away.

6. Roller rink#

Prompt
1980s roller rink: turn the person from the reference photo into a candid 1980s roller rink photo. They skate toward the camera laughing, wearing a bright windbreaker, high socks and four-wheel roller skates, with a disco ball, colored spotlights and a blurred crowd behind. 35mm color film, on-camera flash with motion blur in the background. Keep the person exactly as they are: same face, same features, same skin tone, same age, same hairstyle. No text, no logos.
The roller rink prompt on three models: the man laughing on roller skates in a bright windbreaker under a disco ball
Roller rink, left to right: Nano Banana Pro, Nano Banana 2.1, GPT Image 2.5

All three got the windbreaker, the striped socks and the disco ball. The Nano Banana models pulled the camera back and filled the rink with skaters; GPT Image 2.5 came in close, which makes his face easier to recognize.

7. Walkman street photo#

Prompt
1980s Walkman street photo: turn the person from the reference photo into a 1985 street photograph in a big city. They walk down a sunny sidewalk wearing a cassette Walkman with foam headphones, an oversized blazer with rolled sleeves and high-waisted jeans, old cars and storefronts behind. Warm saturated 35mm slide film, natural daylight, candid documentary style. Keep the person exactly as they are: same face, same features, same skin tone, same age, same hairstyle. No text, no logos, no readable signs.
The Walkman street photo prompt on three models: the man in a blazer and jeans with foam headphones walking down a sunny city sidewalk lined with old cars
Walkman street photo, left to right: Nano Banana Pro, Nano Banana 2.1, GPT Image 2.5

The best street scenes came from the Nano Banana models, with period cars and pedestrians, but both filled the storefronts with made-up lettering despite "no readable signs". This is also where Nano Banana 2.1 lost the face: the man in its picture has a beard but reads as someone else. Small faces in wide shots are the hardest case for every model.

8. Studio portrait#

Prompt
1980s studio portrait: turn the person from the reference photo into a 1980s department-store portrait studio photo. Formal posed portrait with a soft smile, wearing a patterned knit sweater over a collared shirt, set against a mottled blue painted canvas backdrop. Even studio lighting with a soft vignette, glossy color print with slight fading. Keep the person exactly as they are: same face, same features, same skin tone, same age, same hairstyle. No text, no logos.
The studio portrait prompt on three models: the man in a patterned knit sweater over a collared shirt in front of a mottled blue backdrop
Studio portrait, left to right: Nano Banana Pro, Nano Banana 2.1, GPT Image 2.5

The quiet one, and the one people actually frame. Nano Banana 2.1 and GPT Image 2.5 are both very close to the selfie here. Nano Banana Pro produced a lovely picture of a younger, clean-shaven man.

Faces up close#

Close-up crops of the man's face in the original selfie and in the yearbook and Walkman results from each model
Top: laser yearbook. Bottom: Walkman street photo. Original on the left

Thumbnails hide most of the drift. Up close:

  • GPT Image 2.5 reads as the same man in every look: same beard, same eyes, same lines around them. It was the most consistent model in this test.
  • Nano Banana 2.1 keeps the beard and most of the features in close-ups. In wide shots, where the face is small, it can turn into a lookalike (bottom row).
  • Nano Banana Pro keeps his eyes and nose but removes the beard and many years in most looks. Since the prompt asked for "same age", this is the model treating the era words ("high school", "prom", "aerobics video") as stronger than the identity line.

Tips for writing your own 80s prompts#

The first two tips come straight from this test; the others are fixes for the problems above that we did not re-run, so try one image before a batch.

  • Write the clothes for the person in the photo. Models dress the reference person in whatever you describe, prom dress included. If one prompt list is for a mixed group, keep the clothes neutral (a sweater, a denim jacket, a windbreaker).
  • End every prompt with an identity line. It worked on GPT Image 2.5 and mostly on Nano Banana 2.1.
  • Say the facial hair out loud. If Nano Banana Pro shaves you, add "keep the beard exactly as it is" to the identity line.
  • Soften the age words. "1986 high school yearbook portrait" pulls toward a teenager. "A 1986 yearbook-style portrait of an adult" leaves less room.
  • Get closer for likeness. Faces held up best in head-and-shoulders shots; wide street and rink scenes drifted most.
  • "No text" is a request, not a guarantee. Signs, banners and arcade art can still come back with lettering. Check before you post.

Note

Only upload photos of people who have agreed to it, and say the picture is AI-made when you share it.

Make the whole group's 80s photos at once#

For an 80s party invite, a class reunion slideshow or a team page, a chat app means one round of upload, paste, wait and save per person per look. With a bulk image generator the whole group is one batch.

In BulkImagen:

  1. Start a New batch and paste the prompts you want into the prompt box, one prompt per line.
  2. Upload everyone's selfies as reference images and choose Per image under How to use the references. Every selfie is paired with every prompt, each person rendered separately.
  3. Pick the model. Going by this test, GPT Image 2.5 keeps faces best at 10 credits per image at 2K; Nano Banana 2.1 is close behind at 6 and does the best crowd scenes.
  4. Check the total and start. The form shows the image count and credits before you spend anything. Ten selfies with the yearbook and studio portrait prompts is 10 × 2 = 20 images, or 200 credits on GPT Image 2.5. Download everything as one ZIP.

Try one person with your two favorite looks first. If someone's face drifts, use a sharper, front-facing selfie of them. For keeping the same person across many images, see consistent character AI images.

FAQ#

Which AI model is best for 80s photos of yourself? In our test, GPT Image 2.5 kept the face, beard and age in all eight looks. Nano Banana 2.1 was close and cheaper. Nano Banana Pro made the most polished pictures but shaved our subject and made him look younger in six of eight. With one run per prompt and one selfie, test your own photo first.

Can I use these prompts in Gemini or ChatGPT? Yes. They are plain language and work with any model that accepts a reference photo. Gemini's image generation runs on the Nano Banana family; ChatGPT's runs on GPT Image.

How do I get the laser background yearbook photo? Use the first prompt above. All three models drew the purple and blue laser backdrop on the first try; the main difference was how well they kept the face.

Why does the AI make me look younger? Words like "high school", "prom" or "yearbook" describe teenagers, and some models let that win over "same age". Describe yourself as an adult in the scene and name features to keep, like a beard or gray hair.

Can I make 80s photos for a group of friends at once? Yes. Upload all the selfies as reference images, use Per image mode, and every person gets every look in one batch.

Todos los artículosPor BulkImagen Team

Sigue leyendo

Ponlo en práctica

Ejecuta tu primer lote en minutos

Escribe un prompt por línea o sube un CSV, elige un modelo y BulkImagen genera todo el conjunto en paralelo. Las cuentas nuevas reciben créditos gratis para probarlo.