The Mr and Mrs Smith AI trend puts you and your partner into the 2005 film's spy couple, most often in its slow, suspicious ballroom tango. On videosi you add two photos, or one photo of you both, and get a 15-second tango or a 12-second dinner standoff, with no prompt to write.
This guide goes deeper than the page FAQ: why the tango, how it differs from in-app tools, why two faces in close-ups are hard, which tier to pick, and how to post it with an AI label.
Key takeaways
- The tango is 15 seconds and 13 credits on Standard 720p; the dinner standoff is 12 seconds and 10 credits.
- There is no free generation, and credit packs start at $8.90 for 10 credits.
- Two faces in close-ups are hard, and this effect has not been test-rendered yet.
- You cannot use the film's stars: you and your partner play the parts.
- Post it with the platform's AI label. The guns are never fired.
In this guide
- Why is the tango the scene everyone remakes?
- How is TikTok's AI Cast different from a web tool like videosi?
- Why are two-person close-ups so hard to swap faces on?
- Which model tier should I pick?
- How should I shoot the two photos?
- Ballroom tango or dinner standoff: which scene should I choose?
- Can I put the film's stars in it?
- What about the sound?
- Do I need an AI label when I post it?
- Is the gun content a problem?
- What should I expect from the first result?
Why is the tango the scene everyone remakes?
The tango works because it is a scene about two people who cannot look away from each other. In the 2005 film, a married couple discover they are rival assassins hired to kill each other. The ballroom dance compresses that idea into one image: close embrace, steady eye contact, and a hand that may be checking for a weapon.
It also suits a short video. Tango is a partner dance. According to the Wikipedia article on tango, the Argentine and Uruguayan styles use a close embrace with the upper bodies in contact, while ballroom tango is more staccato and uses head snaps. That mix of slow gliding and sharp stops gives a 15-second clip a clear beginning, middle and end without any dialogue.
On videosi the tango runs through four beats: mid-dance on a gold ballroom floor, tight cheek-to-cheek close-ups, a wide dip, and a hand sliding to a thigh holster. The holster is where it stops. The guns are never fired and there is no blood.
How is TikTok's AI Cast different from a web tool like videosi?
Many of the November 2026 tango posts we saw were tagged as made with TikTok's AI Cast feature or with Dreamina. Those are tools inside apps, and apps change them often, so check what they offer today. videosi is independent and not affiliated with TikTok, CapCut or Dreamina.
The practical difference is where the work happens. In an app you usually work inside its template and its editing flow. On the Mr and Mrs Smith AI page you pick one of two scenes, add two photos, choose a model tier and download an MP4. The prompt and the motion reference are built in, so you write nothing. The clip also carries the scene's original sound.
Why are two-person close-ups so hard to swap faces on?
A close-up that fills the frame leaves the model no room to hide. Every pore, glasses frame and hairline is on screen, and in a tango the two faces sit an inch apart. The model has to keep two identities apart while they overlap, move and turn.
There are two typical failures. The faces drift back toward the couple in the reference clip, or the two likenesses blend into each other. For this reason the recipe ties each person to their own photo and names what must go from the original couple: glasses, earrings and hairstyles. Anything left unnamed tends to stay.
We want to be straight about the limit. This effect has not been test-rendered yet, and two faces in close-ups are hard for the Standard model. Likeness is not guaranteed in a 15-second clip. A failed, timed-out or blocked render returns its credits automatically, but a render that finishes and looks only partly like you is still possible.
Which model tier should I pick?
Start from the price list. Credit packs start at $8.90 for 10 credits, and plans start at $29.90 a month. The 15-second tango on the Standard model at 720p is 13 credits, and the 12-second dinner standoff is 10. There is no free generation, so every try costs credits.
Standard is the cheapest way to see whether your photos work. Because the face-to-face section is the hardest part, a higher tier is the better pick when likeness matters more than price, since higher tiers hold faces better. The tiers run on ByteDance's Seedance family, and the exact credit cost for each tier and resolution is shown before you press Generate.
A sensible plan is to run Standard once with your best photos. If the faces hold, keep the result. If they drift in the close-ups, move up one tier rather than retrying the same settings.
How should I shoot the two photos?
Photos matter more than anything else you control. Use a sharp, evenly lit, front-facing shot for each person. Show head and shoulders, because clothes come from the photo and a tight face crop gives the model no outfit to work from. Smart clothes suit the scene best.
Leave sunglasses and hats off. Anything covering the eyes makes the likeness drift during the stare. Photo 1 plays Mr. Smith in the suit and photo 2 plays Mrs. Smith in the dress, and you can swap them however you like.
If you only have a single photo of both of you, that works too. Stand with Mr. Smith on the left and Mrs. Smith on the right, keep both faces clear and side by side, and keep everyone else out of the frame. Both people must be adults who agreed to be in the video.
Ballroom tango or dinner standoff: which scene should I choose?
Choose the tango for the recognisable moment. It is 15 seconds, it has the close-ups and the dip, and it is the scene most remakes use. Choose the dinner standoff if you want a different feel: it is 12 seconds, starts as a dinner party and turns into a standoff by the fireplace. It is also cheaper, at 10 credits on Standard 720p against 13.
The dinner scene has fewer long face-to-face stares, so it may be kinder to a first attempt. The tango is the stronger choice if your photos are crisp and you want the version viewers expect. Both scenes leave the other people in the room as they are in the reference clip: only the two leads are replaced.
Can I put the film's stars in it?
No. Do not upload photos of the original cast or stills from the film: video models usually block famous faces, and the render then fails. The point of the effect is that you and your partner play the parts.
The page is a fan-made scene in the style of the film and is not connected with the film, its studio or its cast. If you want background on the film itself, the Wikipedia article on Mr. & Mrs. Smith (2005) covers the cast, the plot and the music.
What about the sound?
There is no single sound for this trend. Tango posts mostly use the creators' own original sounds, and some remakes use Mondo Bongo by Joe Strummer and The Mescaleros, a song from the film. videosi does not pick a trending sound for you: each scene comes with the original sound of its reference clip already on the MP4.
If you want a particular song, add it in the app where you post. Check that you have the right to use it on that platform.
Do I need an AI label when I post it?
Yes, use one. The clip shows real-looking people doing something they did not do, which is exactly the sort of content platforms ask you to disclose. YouTube's help pages say creators must disclose AI-generated or altered content that seems realistic, and you do this with the AI use setting in YouTube Studio during upload.
Do the same on TikTok and Reels by switching on the platform's AI-generated content setting when you post. Rules change, so read the current community guidelines of the platform you use. A label also keeps the joke honest: the viewer knows the two spies are you.
Is the gun content a problem?
The guns in these clips are props in a staged scene. They are never fired, with no muzzle flash and no blood, and this is not a romantic or kissing scene. That is how the scene is built, but it is not a guarantee about any platform's moderation. Check the current rules before you post and keep the label on.
What should I expect from the first result?
Expect a few minutes of waiting and a result that is close but not perfect. Two-person scenes with many cuts often take more than one attempt, and each attempt costs credits. Check the close-ups first, because that is where the faces slip. If they hold, the rest of the clip usually does too.
If a render fails, times out or is blocked, the credits come back automatically. If it finishes and you simply dislike it, you can try again with different photos or a higher tier. The trend is still emerging, so there is no rush.
Ready to try it? Open the Mr and Mrs Smith AI page, pick a scene and add your photos.



