Short answer: open the I Love The Smiths AI Video Generator, upload a photo of the person who wears the headphones, then a photo of the person who says the line, and press Generate. You get an 11-second 720p MP4 of the two of you in a wood-panelled elevator: one says "I love the Smiths", says it again, hums along and walks out, and the other is left frozen. The voices and elevator sounds are already in it. There is no film clip to upload, no filming and no prompt to write. With only one photo, the AI invents the person who says the line to you.
New to the trend? Read what the I Love The Smiths AI trend is first.
What you need
- Two clear photos, one per person. One photo works too.
- A phone or computer with a browser. There is nothing to install.
- Credits for one clip. The price is shown on the Generate button before you spend anything, and you can buy credits without creating an account.
- A TikTok or Instagram account if you want to post it, and to add the song from the app's library.
How to make the I Love The Smiths video: three ways
People make this clip in three different ways, and each one suits a different person.
| Film it yourself | CapCut template | Two-photo AI generator | |
|---|---|---|---|
| What you need | An elevator, a friend and a third person to film | The app and clips or photos of you | Two photos, or one |
| The elevator | You need a real one | Comes from the template | The AI makes it |
| The lines | You say them yourselves | Usually the film's audio over your clip | Spoken in the clip |
| Risk | People waiting for the lift | The template can disappear | The AI can blur a word or swap who speaks |
| Time | An afternoon | Minutes | A few minutes |
If you have a willing friend and an empty lift, filming it is the most fun version. If you don't, two photos are enough.
Step 1: Choose the photos
The photos decide whether the two people in the elevator still look like you.
- One adult per photo, face toward the camera, sharp and in good light.
- Chest up or waist up. The elevator is a close shot, so faces matter more than shoes.
- A clear mouth helps. The person who says the line needs a visible mouth for the lip sync to look right; a closed-mouth smile is fine.
- Office or everyday clothes look best. Your clothes go into the clip, and a cardigan or shirt suits an office elevator.
- Nothing covering the face. No sunglasses, no low hat and no heavy beauty filter. Real headphones in the photo are fine but not needed; the AI adds them.
- Two separate photos, not one photo of you together, so the AI can tell you apart.
- Any background is fine. The elevator comes from the clip.
Use only your own photos or photos of adults who agreed. No children, celebrities or strangers.
Step 2: Decide who goes where
On the generator page, pick a version. You can do this without an account.
- Two photos: the first upload wears the headphones and stays in the elevator. The second upload says "I love the Smiths", says it again, and walks out. In the film the person who walks out is the cool one, so cast accordingly.
- One photo: you wear the headphones, and the AI invents a stranger who says the line to you. Good for a "this happened to me" post.
Both versions cost the same.
Step 3: Keep the trend settings and generate
The defaults already match the trend: 11 seconds, 720p, vertical 9:16. Eleven seconds is enough for every line plus the exit. 15 seconds costs more but gives the frozen ending more time. Under Advanced you can switch to 480p, which costs less, or to 16:9 if you want it to look like the film.
Press Generate. The button shows the price first; you sign in or buy credits only at this step. The clip is ready in a few minutes and lands in your library, so you can close the tab and come back. If a clip fails to generate, the credits come back automatically.
Step 4: Download and post
Download the MP4 and post it on TikTok, Instagram Reels or YouTube Shorts.
- Sound. The clip already has the voices, a faint guitar tune from the headphones, the elevator hum and the bell. The song in the film is "There Is a Light That Never Goes Out" by The Smiths. It belongs to its owners, so it is not in your file; if you want it, add it from the app's sound library at a low volume under the voices.
- Caption. "I love the Smiths" is what people search for. Other ideas: "POV: you have good taste in music", "me too", or tag the friend who walked out.
- Hashtags people use: #ilovethesmiths, #500daysofsummer, #elevatorscene.
- AI label. Keep the platform's AI label on, since the clip is made by AI.
The scene, line by line
This is what the generator asks for, so you know what to expect:
- The music. Photo 1 stands in the elevator with over-ear headphones on, nodding slightly to a faint tune.
- "I love the Smiths." Photo 2 glances at the headphones, turns with a small smile and says it.
- "Sorry?" Photo 1 slides one side of the headphones off.
- "I said, I love the Smiths." Louder and sweeter. Photo 2 hums along and adds, "You have good taste in music."
- The exit. The bell dings, the doors open and photo 2 walks out without looking back. Photo 1 stays frozen and whispers, "Wow."
There are no captions in the clip, and no logos on the walls.
Three prompts you can copy
The generator writes the prompt for you. If you would rather try the scene in your own video tool, these are written from scratch for this guide. Change anything you like.
1. The elevator (two reference photos)
Keep both people exactly as in their photos: same faces, hair, skin tones and clothes. A small old-fashioned office elevator with warm wood panelling and brass buttons, soft overhead light, one continuous shot from a fixed camera facing them. The person from photo 1 wears over-ear headphones with faint guitar music leaking out. No text, no logos.
2. The lines (add to prompt 1)
The person from photo 2 turns with a small smile and says "I love the Smiths." The person from photo 1 slides a headphone off: "Sorry?" Photo 2, louder: "I said, I love the Smiths," hums the tune and says "You have good taste in music." The bell dings, the doors open and photo 2 walks out. Photo 1 stays frozen and whispers "Wow." Lips in sync, each line said once.
3. Swap the taste
Same scene, but photo 2 points at a vinyl record under photo 1's arm and says "I love this album," then "I said, I love this album." Use it for any band, film or game you two share.
Common problems and fixes
- It doesn't look like us: a face was small, blurry or turned away. Use sharper, brighter photos with faces toward the camera.
- The wrong person speaks: check the upload order. The headphones go on the first photo, the line on the second.
- The two people look alike: use two separate photos, each showing only one person.
- A line is blurred or cut off: keep 11 seconds or more, or run it again. Speech is the newest thing AI video does, and short clips squeeze it.
- The second person doesn't leave: use 15 seconds so the exit has room, or run it again.
- Words appear on the screen or walls: we ask for no text, but AI sometimes draws some. Run it again rather than posting it.
Refunds cover clips that fail to generate, not a take you don't like, so it is worth starting with good photos.
How much does it cost?
The price depends on the length and resolution and is always shown on the Generate button before you start. Two photos and One photo cost the same. Credits work on every app on Vuddex, and credit packs are one-time payments, no subscription.
FAQ
Is this the film with our faces on it? No. The AI makes a new elevator scene from your photos. No footage from 500 Days of Summer goes in and the actors are not in it.
Does it include the Smiths song? No. Add "There Is a Light That Never Goes Out" from your app's sound library if you want it.
Do I need a second person? No. With one photo, the AI invents the person who says the line.
Can I use my pet? This app is written for people talking, and we have not tried pets on it. For a pet clip, try the Dog Making Coffee AI Video Generator.
Ready? Make yours with the I Love The Smiths AI video generator.