The ChatGPT sweater trend, also called the AI sweater trend or sweater portrait, turns one selfie into a black-and-white studio portrait where the only colour is a chunky-knit sweater, with the sleeves pulled up over your mouth. On videosi you upload one photo, pick a colour and get a vertical 2:3 image in about a minute for 1 credit, with no prompt to write.
This guide goes deeper than the FAQ on the trend page. It covers colour choice, writing the prompt yourself, why the pose works, beards and glasses, the Christmas window, a pet version and how to post.
Key takeaways
- The ChatGPT sweater trend is a November 2026 photo trend; the TikTok posts we checked are dated 11 to 14 November 2026.
- Choose a saturated colour far from grey; videosi offers Christmas red and Turquoise.
- A good source photo is sharp, front-facing and well lit, with your eyes clearly visible.
- It costs 1 credit per picture, and the result is a still image with no sound.
- The trend is about a week old and tied to Christmas, so make yours soon.
In this guide
- Which sweater colour reads best on a black-and-white photo?
- Can I write the prompt myself in ChatGPT without copying a long one?
- Why does the sleeves-over-the-mouth pose work so well?
- Does the sweater portrait work for men and for beards?
- What about glasses, hats and hair?
- What makes a good source photo for the sweater portrait?
- Is this just a Christmas trend, and when will it fade?
- Can I make a pet version?
- How do I post the picture, and does it come with sound?
- What does it really cost to get a good result?
Which sweater colour reads best on a black-and-white photo?
Pick a colour that is saturated and far from grey. On a greyscale portrait the sweater is the only large patch of colour, so a strong, clean hue does all the work. The videosi page offers two: Christmas red and Turquoise.
Red is the safe festive choice. It is warm, it sits next to the greyscale skin tones without clashing, and it suits the Christmas season the trend arrived in. Turquoise is the cooler alternative and stands out more in a feed full of festive red, which can help your post look different at a glance.
Pastels and beige are the weak options. They sit close to grey, so the one pop of colour turns into a faint tint and the whole idea of the picture is lost. If you want to compare, making both colours costs 2 credits, which is cheap enough to test.
Can I write the prompt myself in ChatGPT without copying a long one?
Yes, and a short version works if it covers the right points. The original trick is a text prompt you paste into ChatGPT or Gemini together with your own photo. You do not need the exact long text from a caption, but you do need to say the things the model will otherwise guess wrong.
Four parts matter. First, a keep-my-face rule, so you get your own face and not a generic model face. Second, black and white everywhere except one chunky-knit sweater and your natural eye colour. Third, the pose: both hands tucked into long cuffs and pulled up under the nose. Fourth, one picture only, with no text or logos.
Skip any of these and the usual problems appear. We publish the approach and the common mistakes on the trend page, not the full prompt. If you would rather not tune wording at all, videosi runs this for you from one photo.
Why does the sleeves-over-the-mouth pose work so well?
It turns the eyes into the whole portrait. With the lower face hidden behind knitted cuffs, the only expression left is in the eyes, and the sweater frames everything below them. That is why the look reads as a high-fashion editorial shot even from an ordinary selfie.
It also helps the AI. Hiding the mouth removes one of the places where likeness most often drifts, and tucking both hands into the cuffs avoids the extra or odd fingers that image models are known for. The brief for the picture asks for five fingers per hand anyway, but a hand you cannot see cannot go wrong.
The catch is that the model must be told exactly where the cuffs end. If the pose is not described, the sleeves tend to land beside the face and the covered mouth, the signature of the trend, disappears.
Does the sweater portrait work for men and for beards?
Yes. The look works for men, and a beard or moustache can stay visible above the cuffs. The sleeves are pulled up under the nose, so facial hair is partly covered, and what shows is the upper part of the face.
For best results use a sharp, front-facing photo with even light. A heavy beard against a dark background can merge into one dark shape, so a plain, lighter background in your source photo gives the model a cleaner edge to work with.
Colour choice matters a little more here too. A bold red or turquoise sweater against a dark beard gives strong contrast, which is exactly what a one-colour picture needs.
What about glasses, hats and hair?
Hair and age are meant to stay as they are: your features, hairstyle and age are kept, and only the sweater, the black-and-white treatment and the pose are new. Glasses and hats are the part that can change the result, because they cover parts of the face the model uses to match you.
If a first try changes your glasses, try again with the same photo or with a version where your eyes are clearly visible through the lenses. Avoid strong reflections on the lenses and avoid wide hat brims that throw a shadow across the eyes. In this trend the eyes carry the picture, so they should be easy to see in your source photo.
What makes a good source photo for the sweater portrait?
A sharp, front-facing, well-lit selfie with your face in the middle. A dark, blurry or side-on photo makes the likeness drift.
Showing your hands near your chin helps, because the pose ends with both hands in the cuffs close to your face. Keep other people out of the frame: the result must show you as the only person, with no extra hands, faces or background people, and a crowded source photo makes that harder.
You do not need a studio. A window as the light source and a plain wall behind you are enough, and the picture is turned to black and white anyway, so the colour of your own clothes does not matter.
Is this just a Christmas trend, and when will it fade?
It is tied to the season. As of 17 November 2026 the ChatGPT sweater trend is still emerging: the TikTok posts we checked are dated 11 to 14 November 2026, and an earlier start around 9 November is unconfirmed. The trend is only about a week old, so the format may still change.
That points to a short window. Seasonal photo trends tend to peak while the holiday is close and then fade, so our estimate (not a confirmed fact) is that the sweater portrait fades after the holidays. If you want the picture for a holiday card, a profile photo or a post, make it now rather than later.
The knit look itself is not strictly Christmas, so a turquoise version can outlast the red one in your gallery. Still, plan for this to be a few weeks of attention, not a permanent format.
Can I make a pet version?
Not on videosi. The look is built around a person: the face, the cuffs pulled up under the nose and the eye colour all assume a human portrait, so a pet photo is not what the page is designed for.
If you want a pet to join the fun, the practical route is a separate picture. Make your own portrait here and place it next to a normal photo of your pet in a carousel. Our related pages cover other portrait looks, including doodle portrait AI and Christmas profile pictures.
How do I post the picture, and does it come with sound?
The result is a still image with no sound, vertical at 2:3. We know of no official TikTok filter for this trend, so you post the picture yourself: as a photo post, or as a slideshow with a few results in a row.
Most TikTok posts we checked are either a result slideshow or a tutorial. If you want sound, add your own music in TikTok or CapCut. There is no single confirmed song for the trend; the posts we checked used different tracks, such as Christmas songs and original sounds.
A simple format that fits the look: your original selfie first, then the black-and-white portrait, then a second colour. Because the picture has no text, no logo and no watermark, you can add your own caption or sticker without clashing with anything.
What does it really cost to get a good result?
One sweater portrait is 1 credit and a failed or blocked generation returns its credit automatically. If the first picture is not right, a retry costs 1 more credit. A pack of 10 credits is $8.90 on the pricing page, so a few tries in two colours stays small.
Most of the cost is avoidable by fixing the photo first. Sharp, front-facing and well lit gets you a good result more often than any other change. After that, test both colours once and keep the one that suits your skin tone and your feed.
Try it
Ready to make yours? Open the ChatGPT sweater trend generator, upload one clear selfie and pick your colour. It is 1 credit, and nothing needs to be pasted anywhere.



