A Lego couple figures AI photo turns one photo of each person into a picture of the two of you as toy-brick minifigures, keeping your own hair, glasses and outfits. On videosi you upload 2 photos, pick a look and download one vertical 2:3 still image, with no prompt to copy.
This guide goes deeper than the FAQ on the trend page. It covers how to shoot the photos, why the two figures sometimes swap features, how a chat tool compares, how to post it for Valentine's Day, labels and trademarks, and what the picture costs.
Key takeaways
- A Lego couple figures AI photo is a still image of two people as brick minifigures, not a video.
- One clear standing photo per person, with outfit, glasses and hats visible, gives the best match.
- Hair and glasses get mixed when the model is not told which photo belongs to which figure.
- videosi is not affiliated with the LEGO Group and adds no logo or text to the image.
- One picture costs 1 credit; packs start at $8.90 for 10 credits, plans at $29.90 a month.
In this guide
- How should I shoot the two photos for the best match?
- Why does one figure sometimes get the other person's hair or glasses?
- What is the difference between doing it in ChatGPT and using videosi?
- How do I choose between the three looks?
- How do I post it for Valentine's Day or an anniversary?
- Do I need to label the post as AI?
- Is it a LEGO product, and is there a copyright or trademark problem?
- Can I make it on my phone?
- Which details disappear, and what can I do about them?
- How much does one picture cost, and what happens if it fails?
How should I shoot the two photos for the best match?
Take one photo per person, standing, in even front light, with the outfit visible from the waist up or full length. The model copies hair, skin tone, glasses, clothes and shoes from the photo, so anything it cannot see gets invented. A face-only crop leaves the outfit to chance.
Keep the things that make a person recognisable: glasses, a beard, a cap, a big hairstyle, a bag. A minifigure face is just two dots and a smile, so these accessories do most of the work of saying "this is you". Remove sunglasses and heavy filters, because they hide the features the figure is built from.
Use a plain wall or a quiet room behind you. A busy background makes it harder to separate a person from what is around them, and bits of the background can leak onto the figure. Crop group shots first, since only the most prominent person in a photo is used.
Why does one figure sometimes get the other person's hair or glasses?
This is the classic failure of two-person prompts. When a model is told to draw two people but not told which photo belongs to which figure, it blends them. Both figures end up with the same hairstyle or one pair of glasses, and the pair looks like two copies of an average person.
The fix is an identity rule for each photo: Image 1 is the first person and only the first figure copies from it, Image 2 is the second person and only the second figure copies from it. The page keeps that rule built in, so you do not write it yourself. If you try it by hand in a chat tool, name the photos explicitly and say never to mix or swap features.
A second cause is two people who look alike in the photos, for example both with dark hair and a dark top. If a result mixes them, try photos where each person has a clearly different top or accessory.
What is the difference between doing it in ChatGPT and using videosi?
Both routes use the same kind of image model, so the difference is mostly in effort and control. In a chat tool you upload two photos, paste a long prompt that you found in a tutorial, and then fix problems by chatting back and forth. Results depend on the exact wording, and you may need several tries.
On videosi you upload one photo of each person, pick one of 3 looks and press Generate. The prompt is private and already handles the identity rule, the toy style, the pose and the "no text or logo" rule. A generation costs 1 credit, and a failed or blocked one returns its credit automatically.
If you already pay for a chat assistant and enjoy iterating, doing it by hand is fine. If you want a repeatable result in about a minute from a phone, the Lego couple figures AI photo page removes the prompt work.
How do I choose between the three looks?
The clean studio pair puts the two figures full length on a white backdrop with a soft shadow. It is the safest choice for a profile picture, a message or a caption post, because nothing competes with the figures.
The miniature scene builds a small brick diorama of the place in your first photo, such as a table, a room or a street, at the figures' eye level. Pick it when the setting matters, for example a first-date restaurant or the home you share. Because it reads the place from your photo, send a first photo where the background is visible.
The gift display case puts the pair in a clear case on a small black baseplate with brick flowers and a heart. It suits a card, a printed gift or a Valentine's post. The case carries no sign or writing, so add your own caption. All 3 looks cost the same 1 credit, so you can try more than one.
How do I post it for Valentine's Day or an anniversary?
The result is a single vertical 2:3 still, so post it as a photo, a carousel slide or a story, not as a video. Captions that people already search and use are "us as Lego figures" and "Lego couple", and a date or a short line about the story behind the pair works better than a generic caption.
A good pattern is two slides: your real photo first and the minifigure pair second, so followers see the before and after. If you want motion, add your own music in the app you post from. The image itself has no sound.
Clips of this idea have been posted since mid 2025 and new ones were still appearing in November 2026, so it is a familiar format rather than a one-off. For Valentine's Day, posting a few days before 14 February gives it time to be seen.
Do I need to label the post as AI?
Most social apps ask you to label content that was made or heavily edited with AI, especially when it shows real people. A toy-style picture of you and your partner is clearly stylised, but the safest habit is to use the platform's AI label or to add "AI" in the caption. The exact rules differ by app and change over time, so check the current settings in the app you post to.
Labelling costs nothing and avoids a removal or a reduced reach later. It also helps friends understand that the picture is a stylised portrait, not a photo of physical toys you bought.
Is it a LEGO product, and is there a copyright or trademark problem?
No. videosi is not made by, endorsed by or affiliated with the LEGO Group, and the image carries no brand logo, wordmark or text. The style is brick-toy inspired: a cylindrical head, a stud on top, C-shaped hands, a blocky body. LEGO is a trademark of its owner, so do not add the logo to the picture or present it as an official product.
For personal posts, gifts and prints for the two people in the image this is the common use. If you plan to sell the pictures or use them in an advertisement, check the trademark and image rules that apply to you, and keep any brand names out of the artwork and the listing text.
Only upload photos of adults who agree to be in the picture, two people per image. Do not upload a photo of a child, and do not use the tool to mislead anyone about a real person.
Can I make it on my phone?
Yes. The page works in a phone browser, so there is nothing to install. Open it, upload a photo of each person from your gallery, pick a look and press Generate. A result usually comes back in about a minute, and you save it to your photos from the result screen.
If the two of you are not together, ask your partner to send a clear standing photo, or each of you can take one. If you only have one photo of you both side by side, upload that and the page treats it as the source for both figures; a photo where you stand a little apart works best.
Check that the photo is sharp before you upload. A blurry or dark photo is the most common reason a figure does not look like the person.
Which details disappear, and what can I do about them?
A minifigure is a simplified shape, so fine details can change or vanish: tattoos, freckles, small jewellery, thin glasses frames and complex prints. Hair shape, hair colour, skin tone, glasses, hats and clothing colours carry over best.
If a detail matters, make it big in the photo. A bold necklace or a large logo-free print on a shirt survives, a tiny pendant does not. If a result misses something important, generate again with a photo where that detail is clearer, or pick the studio look, where the figures are shown larger and cleaner.
The image has no text of any kind, so a slogan on a shirt will not be reproduced as readable words.
How much does one picture cost, and what happens if it fails?
One Lego couple figures AI photo costs 1 credit. Credit packs start at $8.90 for 10 credits, and plans start at $29.90 a month. There is no free generation on this page. If a generation fails or is blocked, the credit comes back automatically, so you do not pay for an image you did not get.
Because all 3 looks cost the same, a sensible way to spend credits is to generate the studio pair first, check that both figures look like their people, and then try the scene or the display case once you are happy with the photos.
For the full steps and the three looks, see the Lego couple figures AI photo page. For other two-photo couple ideas, the couple wallpaper AI prompt page makes a phone wallpaper with one eye of each of you, and the bobblehead AI page turns one photo into a toy figure of a single person.



