A 70s AI photo turns one selfie into a portrait of you in the 1970s, with feathered or shaggy hair, flared jeans, a wide-collar shirt and warm film grain. On videosi you skip the prompt: upload one photo, pick a setting and get one vertical 2:3 still image for 1 credit.
This guide goes deeper than the FAQ on the trend page. It covers where the trend came from, the two settings, how hair and clothes change, how to keep your face, which photo to use, how it compares with Gemini and ChatGPT, and how to post the result.
Key takeaways
- The look is a colour 1970s snapshot with rust, mustard and denim tones, fine grain and a soft vignette, not a black-and-white filter.
- The result is one vertical 2:3 still image made with GPT Image 2.5 from OpenAI, not a video, and it has no sound.
- Pick the street with a rust-orange car or the wood-panelled living room, and use one sharp front-facing photo to keep the face yours.
- Your face, age and skin tone come from your photo; only the hair, clothes and setting change.
- Each image costs 1 credit, and a failed render returns the credit automatically.
In this guide
- What counts as a 70s AI photo?
- Where did the trend come from?
- How do you make one from your own selfie?
- Street or living room: which setting should you pick?
- How does the hair and clothing change by person?
- How do you keep the face looking like you?
- Which photo works best?
- How does this compare with Gemini and ChatGPT?
- Why is the result colour and not black and white?
- How do you post the result?
- What does it cost and what if it fails?
What counts as a 70s AI photo?
A 70s AI photo is a picture of you restyled as if it were a colour snapshot from the 1970s. The look is feathered or shaggy hair with sideburns, wide collars, flared or bell-bottom trousers, platforms or leather boots, and warm earth tones such as rust, mustard, brown and denim blue. The colour is faded like old film, with fine grain, soft glow around lights and a gentle vignette.
It is not the same as a filter. A filter tints your own photo and leaves your clothes and hair as they are. The trend rebuilds you in 1970s hair and clothes inside a scene, which is why a modern selfie in a hoodie can still become a convincing street or living room portrait. The 70s AI photo trend page shows examples of the finished look.
Where did the trend come from?
There is no single original creator we can name. Thai tutorials showing a Gemini prompt for a 1970s style photo were already posted in October 2025, and one of them passed 2 million views. A bigger wave followed in February and March 2026, when posts titled "1970s filter effect AI" and "1970s vintage filter AI" each reached hundreds of thousands of views.
Many of those posts are black-and-white vintage portraits, not disco looks. That matters if you search for the trend: the same words can lead to an old studio portrait or to a colourful street scene. By late October 2026 the trend had cooled, but 1970s AI clips, from CapCut templates to AI cast videos, are still being posted.
How do you make one from your own selfie?
On videosi it takes three steps. Upload one clear photo of yourself, pick a setting, and press Generate. You get one vertical 2:3 still image, usually in about a minute. Your face, age and skin tone come from your photo, and the hair, clothes and film look are added for you.
There is nothing to write. The page fills in the prompt, so you do not paste anything into a chat box. You can start on the trend page and be done in a few minutes, on a phone or a computer, with no app to install.
Street or living room: which setting should you pick?
You get two settings, and they give different posts. In the street setting you lean back against the rear wing of a rust-orange 1970s hatchback on a sunlit city street, full-body, one hand on a hip and the other in a pocket, with old shopfronts and a few people far behind you. It shows off the flared jeans and the boots, so it suits a full-length post.
In the living room setting you sit on an olive-green sofa, waist-up and angled to the camera. Around you are wood-panelled walls, orange floral curtains with sheer drapes, a shag rug, a wooden stereo cabinet and a lava lamp, lit by warm lamp light mixed with window light. It is closer and cosier, and it shows more of your face and shirt. If you cannot choose, make one of each. They look quite different side by side.
How does the hair and clothing change by person?
The styling follows the person in your photo. For a woman, the result is long soft feathered hair parted in the middle or a natural afro, a printed or earth-tone blouse with a wide collar or a wrap top, high-waisted flares, platform shoes and a thin belt. For a man, it is layered shaggy or side-swept hair with sideburns, a patterned wide-collar shirt open at the neck or a ribbed turtleneck, flared trousers and leather boots.
Glasses and facial hair from your photo are kept, and a short moustache is fine if you already have one. Only the hairstyle, clothes and setting change, so people searching for a boys version or a girls version do not need a separate prompt.
How do you keep the face looking like you?
The most common complaint about AI retro photos is that you turn into a more glamorous stranger. The fix is to put the identity rule first and make it strict: the face, age, skin tone, features, build and natural hair colour all come from your photo, and only the hairstyle, clothes and setting change. Short copied prompts often skip this, so the model improves your face without being asked.
Likeness is still not guaranteed. A new hairstyle, sideburns and warm film colour can shift small details. When the result looks like a cousin, try a different photo before changing anything else.
Which photo works best?
Use one front-facing photo in soft daylight with the whole face visible and nobody else in the frame. A shoulders-up or waist-up shot is ideal. Avoid sunglasses, masks, hands near the face and heavy beauty filters, since a filtered selfie gives the model a different face to copy. Clear glasses are fine and are kept.
Group photos do not work well, because the model cannot know which person you mean. Crop the picture down to the one face you want. The clothes in your photo do not matter, because the 1970s outfit is added for you.
How does this compare with Gemini and ChatGPT?
Most 70s AI photos on social media came from a long prompt pasted into Gemini or ChatGPT together with a selfie. You can still do that, and it can work. The usual problems are that the face drifts, the clothes look like a costume, and a plain request for "70s" gives a tinted modern photo.
With videosi you choose only the setting. The identity lock, the 1970s hair and clothing details and the film look are built in, and the image is made with GPT Image 2.5 from OpenAI. The trade-off is that it is not free, while a copied prompt can be tried in Gemini itself under its own limits. videosi is independent and not affiliated with Google, OpenAI, TikTok or CapCut.
Why is the result colour and not black and white?
Many posts called "1970s filter effect AI" are black-and-white vintage portraits, like an old studio shot. That is a different look. The videosi version is a colour snapshot, in the style of warm faded Kodachrome film, with rust and mustard tones, fine grain and a soft vignette.
Everything in the frame belongs to the 1970s. There are no smartphones, no modern cars, no LED screens and no modern signage, and any car is a plain 1970s model with no badge or readable plate. People in the background are adults, small and out of focus. There is no text, logo or watermark on the image.
How do you post the result?
The result is a vertical 2:3 still image, which sits well on an Instagram feed post and fits a Story with slight cropping. It has no sound, so add a disco or soul track yourself if you turn it into a reel. There is no single confirmed sound for this trend, so pick any 1970s track you like.
Because the image looks like a real old snapshot, say that it is AI made when you share it publicly. For a couple or a family, make one image each with the same setting and post them together. The photo is made for one adult per image, and you should only use photos of adults who have agreed.
What does it cost and what if it fails?
Each image costs 1 credit. Credit packs start at $6.99 for 10 credits, and credits never expire, so you can keep them for the next trend. If a render fails or is blocked, the credit is returned automatically, so you only spend credits on images you actually receive.
Making both settings from one photo costs 2 credits in total. You can open the 70s AI photo trend page at any time, because it stays live even after the trend has cooled.
