Full 20-second scene · 1 or 2 photos · Sound on
Upload your photo, or yours and your crush’s. You get the whole 20-second elevator moment, not a 15-second cut: headphones on, “I love the Smiths.” A missed line, “I said, I love the Smiths,” headphones off, the look, “You have good taste in music,” and a line of the song. Your faces, your outfits, the sound already synced.
Make yours
No prompt, no elevator, no camera. The scene, the lines, the timing and the sound are built in. Pick who you play; add the other person’s photo, or keep our AI co-star.
Watch and download on your private order page or from the email link. Photos are deleted within 7 days.
Why the full version
Many AI templates for this trend stop at about 15 seconds, re-voice the lines or require exactly two photos. We follow the whole exchange, beat for beat. And our example is a real result made from the photos next to it, not a TikTok clip.
| I Love the Smiths AI | Typical 15-second AI templates | |
|---|---|---|
| Examples | A real result generated on this site, shown next to the two photos it was made from | Often the original TikTok clip or film footage, not a generated result |
| Length | 20 seconds, the whole exchange | About 15 seconds |
| Lines | All three: “I love the Smiths”, “I said, I love the Smiths”, “You have good taste in music” | Often only the first one or two |
| Ending | The look and a sung line of the song | Often cut before the song |
| Sound | The scene’s own voices and music, synced to the mouths | Sometimes new voices or music |
| Photos | 1 or 2 photos, same price; you pick your role | Usually exactly 2 photos, fixed roles |
| Account | No login, one-time payment | Often sign-up and credit plans |
Based on public template pages in November 2026; details vary by site.
Make the full version →The scene
One fixed camera, one take, two people side by side in an elevator. Every video follows the same scene and the same lines. Only the people change: faces, hair and the colors and style of clothes come from your photos.
Left: lost in a song, headphones on, eyes ahead. Right: hears the music leaking out and turns with a smile.
The first line, said to someone who can’t quite hear it.
Second try. The headphones come off and settle around the neck.
Both turn to each other. Surprise, then a slow smile, while the song keeps playing.
The compliment that starts everything, with a small gesture.
A soft line of the song, a grin, and a glance back. The doors could open any second.
How it works
You only choose the photos and who plays whom. The elevator, the lines, the camera and the sound are already set.
A clear, front-facing photo of one adult per photo, chest-up or waist-up. Pick whether you are the one in headphones or the one who says “I love the Smiths”, and add the other person’s photo if you like.
Tick that you may use the photos, enter the email for delivery and pick a pack. No account or password.
Your video starts as soon as payment is confirmed and keeps going if you close the page. It is usually ready in 5–15 minutes, and we email you when it is done.
Download the 20-second MP4 with sound and post it to TikTok, Reels, Shorts or your story. Add your own caption, like “I said I love the Smiths”, in the app.
What you get
Not a talking photo or a face sticker: the whole scene is re-created with two moving people who speak, turn and react.
We ask the AI to keep the face, hair, skin tone and build from each photo, including when you turn, take off the headphones and look at each other. AI can still drift, so a sharp, front-facing photo matters most.
Upload your photo →The colors and style of your tops come from your photos; details such as necklines, sleeves or layers can change. The elevator, the lighting and the headphones are always the same, and the room in your photo is not used.
See the scene →The video comes back with the scene’s sound in place: the spoken lines and the music, timed to the mouths on screen. No editing, no lip-sync app.
Watch with sound →No account and no public gallery. Your photos are used only for your video and deleted within 7 days; finished videos stay on your private order page for 30 days.
Read the privacy policy →Know before you start
Hair color and length come from your photo, but the AI may style long hair like the scene, for example tied back in a side braid. Without a second photo, the other person is our AI-generated co-star, not an actor from the film. You can’t change the lines, the voices or the song. That keeps every video true to the trend, and means there is no prompt to write.
Photo tips
The camera stays close on two faces for 20 seconds, so the face in your photo is what people will look at.
Ideas
Put the two of you in the scene where it all began for Tom and Summer. With their agreement, it is the most low-key “so, about us” post there is.
For the couple who bonded over a band, a playlist or a shared pair of earbuds. A 20-second gift that needs no explanation.
Pick either role and play opposite our AI co-star. Post the elevator moment without finding a scene partner or an elevator.
Record-shop staff, radio hosts and playlist curators who agreed can star in a light-hearted 16:9 clip. For ads or brand accounts, use licensed music instead of the scene sound.
Format and delivery
Pay once
Each credit makes one new 20-second, 16:9, 720p “I love the Smiths” elevator video from one or two photos, sound included. No subscription, no account.
$12.99 per video
$10.83 per video
$9.10 per video
Questions
It comes from the elevator scene in the 2009 film (500) Days of Summer. Tom is listening to The Smiths on his headphones; his new co-worker Summer hears it and says “I love the Smiths.” He doesn’t catch it, so she says it again, and adds “You have good taste in music.” On TikTok and Reels, people now put themselves, their crush or their partner into that moment.
The full version. Your video runs about 20 seconds and follows the whole exchange in order: the headphones, “I love the Smiths”, “I said, I love the Smiths”, the headphones coming off, the look, “You have good taste in music” and a sung line of the song. Many AI templates for this trend are about 15 seconds and stop earlier.
Yes. The first example on the home page was generated by this site’s own pipeline from the two photos shown next to it (both AI-generated fictional adults), not taken from TikTok or the film. You can compare the faces and tops with the photos yourself. Some sites for this trend show the original TikTok clip or film footage as their example instead of a generated result.
Upload a clear photo of yourself, pick whether you are the one in headphones or the one who says the line, optionally add the other person’s photo, confirm you may use the photos, enter your email and pay. You get a private link to watch and download the video. No filming, no elevator and no prompt.
Yes. Upload your photo, choose your role, then add their photo: they play the other person. It costs the same one credit as one photo. Use two separate photos with one person each, and only with their agreement; a single photo of both of you doesn’t work.
Yes. Pick your role and leave the second photo empty: the other person is our AI-generated co-star, a fictional adult, not an actor from the film.
You choose. “The one in headphones” stands on the left, misses the first line and takes the headphones off. “The one who says it” stands on the right and says “I love the Smiths.” With two photos, your photo takes the role you picked and the second photo takes the other. Either role works for any adult, whatever their gender.
No film footage and no actors: the whole scene is AI-generated with the people from your photos. The video comes with the scene’s own sound, the spoken lines and music, already synced. It is a re-creation inspired by the film, not a clip from it.
No. The lines, voices, timing, camera and elevator are the same in every video so it matches the trend. The voices do not take on the sound of the people in your photos.
No app and no account. It works in the browser on iPhone, Android or a computer: pick photos from your camera roll, pay, and download a normal MP4 you can post to TikTok, Instagram Reels or YouTube Shorts.
The video is landscape 16:9, framed like the film, so both faces are in view. TikTok, Reels and Shorts all accept landscape videos; you can also place it on a vertical background with the app’s own editor and add a caption above it.
One adult per photo, face clear and looking at the camera, chest-up or waist-up, in even light without heavy filters. Group shots, sunglasses, side profiles and blurry images make the result look less like the real person.
We ask the AI to keep each face, hair, skin tone and build in every shot, and to follow the scene’s mouth movements so the lines match. AI can still change small details, so watch the result with sound. Sharp, front-facing photos help most.
Your hair, glasses and the colors and style of your tops come from your photos; details such as necklines or sleeves can change. The listener always gets the black wired headphones. The elevator, camera and sound are fixed; you can’t choose another background or song.
About 20 seconds, landscape 16:9, 720p MP4 with sound and no watermark. That is the whole exchange in one take, from the headphones to the song.
Packs are one-time USD payments: 1 video $12.99, 3 videos $32.49, 10 videos $90.99. One or two photos cost the same. There is no subscription and no free personal preview; you can watch the scene before paying. Every new video uses one credit.
You pay once on a secure checkout page run by Dodo Payments, our payment provider. Your card details go to them, never to us. Prices are in USD; any applicable tax is shown before you pay, and your bank may convert the currency.
Usually 5–15 minutes, depending on the video service’s queue. You can close the page right after paying: your order page updates by itself and we email you when the video is ready, or if it failed and your credit came back. If it takes much longer than that, email support with your order reference; we reply within 1 business day. Your credit stays reserved, so don’t pay again.
Through the private link in your order emails, which works on any device for 30 days. The browser you paid with also lists your recent orders under My orders. If the link has expired, email support from the address you used at checkout and we’ll restore access to your unused credits.
Use your own photo or photos of adults who agreed to the video. Don’t upload celebrities, public figures, minors or anyone without permission. The scene is made for two humans, so pets, cartoons and group photos are not supported.
No. I Love the Smiths AI is an independent website. It is not affiliated with The Smiths, the makers of (500) Days of Summer, TikTok, CapCut or ByteDance. The name describes the trend people search for.
You can post it on your own accounts. The music in the scene belongs to its rights holders and isn’t licensed by us, so for ads, sponsored posts or brand accounts, replace it with music you or the platform have licensed. Only feature people who agreed to be in the video, and label it as AI-made where a platform asks.
If the video service reports a failure, the credit for that attempt goes back to your order automatically, once. You’ll see it on your order page and get an email, and you can try again with the same or other photos at no extra cost. A returned credit is not a cash refund; for refunds, see the refund policy or contact support.
We delete your photos from our storage within 7 days and keep finished videos on your private order page for 30 days. Nothing is published to a gallery. The video provider that makes your video processes the photos under its own policies, described in our privacy policy.
Guides
Step by step: pick the photos and roles, make the video, caption it and post it with sound.
02The film, the song, the lines and why the elevator moment keeps coming back.
03Three ways to make it, what each needs and which result to expect.
One photo is enough; two make it a story. No filming, no lip-sync, no editing.
Make my I Love the Smiths video →