Skip to main content

AI APT MV maker: one of you on drums, the other on the mic

Two photos in. A pink studio, a drum kit, a chrome mic and a shared last line out.

An AI APT MV puts two people into the pink-lit studio look of the 2024 duet video that turned into a trend: one behind a drum kit, one at a vintage mic, both hamming it up. You add a photo for each person and get a 10-second vertical clip with both faces in it.

1Add two photos(One clear portrait per person)
Quick start
Drummer (left)
Singer (right)

One clear, front-facing photo for each person. Each keeps their own face, hair and skin tone; the booth, outfits and mic come from the template.

No nudity, minors or other people's photos without consent — every upload is screened.

Kept from the original: A pink-lit studio, a drummer on the left and a singer on the right, one shared chorus
2Make it about you (optional)

Add names or what it's for and we'll write the short lines they sing, in the language you type. Leave the topic empty and it's this site's language. Leave everything empty for the template's own lines.

Pick an occasion
3Generate

A 5-second first verse: the left performer raps. The full 10 s duo comes after.

This video: 13 credits · about 23 videos a month for $49.90 on Pro, or $500 a year

No subscription needed · Usually ready in 10-15 minutes · Failed renders are refunded

Example output10s · 9:16

Pink Studio Drum Duo

Choreography locked · Facial reference enabled

MiniMax H3
Format: 9:16 Vertical HDMiniMax H3

The first photo is cast as the drummer on the left, the second as the singer on the right, and the model usually keeps to that. The clip has its own made-up lines and beat, not the original song, so you can add the sound you want when you post.

  • Input: two clear photos, one person each. Photo one is cast as the drummer, photo two as the singer.
  • Output: a 10-second vertical 9:16 video with vocals and a drum beat made for the clip.
  • Price: 13 credits per clip, shown before it renders. The cheapest pack that covers one is the Duo Pack, $29 for 50 credits.

What an AI APT MV is, and where the look comes from

In October 2024 a pop-rock duet came out with a music video shot in a small room washed in pink and magenta light. Two singers traded lines, one of them sat behind a drum kit for part of it, and the hook was named after a Korean drinking game where players stack their hands and call out a floor number. The song went to number one in several countries, and fans started recreating the room and the hand game almost straight away.

The AI APT MV trend takes that look and swaps the two performers for any pair: best friends, a couple, two coworkers, a parent and an adult child. What makes a clip read as the trend is the pink studio, the drum kit next to a microphone, and two people clearly having fun with each other. The exact recording matters less than you would think, which is why the audio can be added later, when you post.

Our version builds the studio, the drums, the mic and the two roles. It does not use the original recording, the lyrics, the footage or the faces of the two artists, and it is not connected to them or their labels. What you get is a clip of the two people in your photos, in that kind of room.

AI APT music videos we rendered ourselves

Each clip is the full 10-second render from the Pink Studio Drum Duo template on two AI-generated portraits, with no editing. None of them shows a real person. Watch the clip first, then read what went in and what drifted.

Duo 1: a black bob on drums, a curly-haired singerIn: Two AI-generated portraits: a woman with a short black bob in a denim jacket (photo 1), and a bearded man with short curly hair in a white T-shirt (photo 2).Out: She plays a silver kit on the left while he sings into a chrome mic on the right, under pink light with a soft halo behind them. Around the middle he turns to her for the clapping game, then goes back to the mic for the last line.A full render from the Pink Studio Drum Duo template. Both faces held and stayed on their own side. He came out wearing a denim jacket over his T-shirt, borrowed from her photo, and the bass drum has a made-up logo on it.
Duo 2: a red beard on drums, a blonde singerIn: Two AI-generated portraits: a man with a ginger beard in a green flannel shirt (photo 1), and a woman with a long blonde ponytail in a black top (photo 2).Out: He drums and grins on the left, she sings on the right. The two trade claps halfway through, and she ends pointing at the camera while he keeps the beat.A full render. Both outfits matched the photos this time. The light drifted from pink to teal for a few seconds in the middle, so the pink room is not guaranteed for the whole clip.
Duo 3: the take where the roles swappedIn: Two AI-generated portraits: a woman with long wavy dark hair in a mustard sweater (photo 1), and a man with black glasses in a grey hoodie (photo 2).Out: The model put him on the drums and her at the mic, the reverse of the photo order. Otherwise the clip runs to script: she sings, the two clap hands in the middle, and she finishes the last line while he plays.A full render. Both faces and both outfits held; only the roles flipped. The video model also refused this pair on the first try and accepted it on the second, and the provider did not bill us for the refusal.

What happens in the clip, second by second

This is the script the template gives the model. Our renders followed it closely, though not to the exact frame, and knowing it helps you decide who should be the drummer.

TimeDrummer (photo 1, left)Singer (photo 2, right)
0 to 4 sCounts in with four hits, looks over and raises an eyebrowGrabs the mic and sings the first line, shoulders moving with the beat
4 to 7 sPuts one stick down and joins the clapping game with a free handSteps off the mic and starts a stacked-hands clapping game with the drummer
7 to 10 sKeeps playing and leans in toward the micBack at the mic; both sing the last line, then point at the camera on the final hit

The photo order is how you ask for these roles. In one of our three renders the model swapped them anyway, as the render notes below record.

Picking the two photos for an APT AI video

Each face is read on its own and then placed in its role, so the pair of photos matters more than either one.

✓ Works

  • One person per photo, face to the camera

    Two separate photos, each a clear head-and-shoulders shot. You do not need a picture of the two of you together.

  • Matching light

    Two photos taken in similar daylight come out as a believable pair under the same pink light. A dark photo next to a bright one gives the model less to work with on the darker face.

  • The livelier person on the mic

    The singer gets the first line and the most screen time. The drummer reacts more than they perform.

✕ Breaks it

  • Sunglasses, hats pulled low, hands on the face

    Hidden features get invented, and the invention can change between the start and the end of the clip.

  • A group photo of three or more people

    The model chooses two of the faces, and not always the ones you meant. A photo of just the two of you works: tap "Use one group photo" under the photo slots, and the person on the left of the photo plays the drums.

  • Celebrities, pets or cartoon characters

    The template is for real people who agreed to be in it. Our photo check refuses photos without a clear person in them, and the confirmation you tick rules out celebrities.

The sound in your AI APT music video

The clip comes with its own audio: a pop-rock drum beat, a guitar riff and two voices singing short lines written for the template. None of it is the original song, and none of it quotes the lyrics. That is on purpose. A clip with the real recording baked in is the kind of file that platforms mute or take down, and it would put a song we have no licence for inside your video.

If you want the actual track, add it where you post. TikTok, Instagram Reels and YouTube Shorts all carry licensed versions of popular songs in their sound libraries, and attaching a sound from the library is how most people in the trend do it. Turn the original clip audio down to zero in the editor first, so the two tracks do not fight.

If you keep our audio, the clip stands on its own as a short duet. The lines are deliberately simple so they still work as background under a caption.

What one AI APT MV costs on AI Crying Trend

Each 10-second clip is 13 credits, whichever pack paid for it. There is no free preview and no sign-up credit, so this table is the whole price list for packs.

DetailCreditsPriceAPT clips it pays for
Single10$9.990
Duo Pack50$293
Crew Pack350$9926
Studio Pack900$19969
Label Pack3000$499230

Every pack is a one-time payment and its credits never expire; monthly plans are on the pricing page. Pressing Generate with too few credits opens the payment window before anything renders, and a render that fails on our side has its credits refunded.

Pairs that make the APT duo trend land

The quiet one on drums

Put the friend who never dances behind the kit. The joke is watching them count in and play along.

Long-distance best friends

Two photos taken in different cities end up in one room. Send it on a birthday instead of a voice note.

Couples

The clapping game in the middle reads as flirting without anyone having to act it out.

A team intro

Two new colleagues in a ten-second duet introduces them better than an email, as long as both said yes first.

When an APT mv AI render comes out wrong

Most bad takes come from the photos, not the template. These fixes are in the order we would try them.

  1. 01

    The wrong person is drumming

    Render again first: in one of our three test clips the model swapped the two roles on its own, and the next render followed the order. If it keeps happening, add the same two photos in the opposite order.

  2. 02

    A face looks soft or generic

    Swap in a sharper, closer photo of that person, taken in daylight, with nothing covering the face.

  3. 03

    Clothes do not match the photo

    Expect this. The photos fix who is in the clip, not what they wear, so outfits often change to suit the studio.

  4. 04

    The render is refused

    The video model runs its own content check, and it sometimes turns down a pair of ordinary photos. One of our three test pairs was refused on the first try. When that happens its credits are refunded; try again, or use different photos.

  5. 05

    The ending feels off

    Two renders of the same pair never match exactly. If the faces are right but the last line is not, rendering again is a reasonable next step.

What this tool will not make

It is not for putting a celebrity, a public figure or anyone who has not agreed into a clip, and it does not reproduce the original music video. Before generating you confirm that both people are you or adults who said yes. It is built for people, so pets and cartoon characters do not work in it. If someone in your clip asks you to take it down, take it down.

Render notes for the three APT clips

What we ran and what went wrong along the way.

ItemDetail
Rendered2 October 2026
Model and modeMiniMax H3, reference to video, two reference photos (drummer left, singer right)
InputsSix AI-generated portraits in three pairs, no real people
Length and editing10 seconds each, full renders, no editing
Render time195 to 265 seconds per clip once the render started
What went wrongIn 1 of 3 clips the model swapped the roles, with the second photo on drums. The provider also refused one ordinary pair on the first try and accepted the same pair on a retry
What the test cost usAbout US$0.60 for six portraits and three renders

Three renders is a small sample, so treat the role swap as a risk, not a rate.

Questions about the AI APT MV

What is the AI APT MV trend?

It is a short video format where two people are placed into the pink-lit studio look of a 2024 pop-rock duet video, one behind a drum kit and one at a microphone, trading lines and goofing around. The hook of that song is named after a Korean party game where players stack their hands and call out a number, which is why so many clips include a hand-stacking moment. In the AI version you upload a photo of each person and an AI model builds the room and the performance around your two faces. Friends, couples and coworkers use it as a ten-second inside joke.

How do I make an AI APT music video with a friend?

Add one photo for the drummer and one for the singer in the tool at the top of this page, then confirm that both people are you or people who agreed, with a parent or guardian agreeing for anyone under 18. Generating asks you to sign in with Google, and your photos are kept while you do. A clip is 13 credits, and if your balance is lower, a payment window opens first and nothing renders until it is paid. The 10-second clip then waits in a shared queue, usually 10 to 15 minutes, and you can leave the page while it runs. It lands in your dashboard as a vertical MP4 you can download and post. There is no prompt to write, because the scene is fixed.

Is the APT AI trend free to make?

Not here. AI Crying Trend has no free tier, no free preview and no sign-up credits. Each 10-second clip is 13 credits, and the button shows that before anything renders. The Single pack (10 credits, $9.99) does not cover one clip; the cheapest that does is the Duo Pack at $29, which pays for 3. Packs are one payment each, never renew, and their credits never expire. If a render fails on our side, its credits are refunded. Whichever site you use, read the price on the generate button before you upload your photos.

Which photo becomes the drummer and which the singer?

The first photo you add is cast as the drummer on the left, and the second as the singer on the right. The singer opens with the first line and gets more screen time, while the drummer counts in, reacts and joins the clapping game, so put the more outgoing person on the mic. The order is an instruction to the model, not a guarantee: in two of our three renders it followed the order, and in the third it put photo two behind the drums. Each face still kept its own likeness. If the roles come out the wrong way round, render again or swap the photos.

Does the video include the original APT song?

No. The clip has a drum beat, a guitar riff and two voices singing short lines written for this template, and none of it is the original recording or its lyrics. We do not have a licence for that song, and a clip with it baked in is likely to be muted or removed by the platform. To post with the real track, open the clip in the TikTok, Instagram or YouTube Shorts editor, turn the original audio down to zero, and pick the song from the platform sound library, which is licensed for use there. That is how most people in the trend add it anyway.

Can I put a celebrity or a K-pop idol in it?

No. The tool is for you and people who have agreed to their face being used. Putting a real public figure into a music video they never made is the kind of clip that platforms label as misleading and remove, and it can break likeness laws in several countries. Before generating you confirm that both people are you or consenting adults. If you want a fan tribute, make it with your own face in the drummer or singer role instead of someone famous.

Why do the clothes in my clip differ from the photos?

Because the two photos set who appears in the clip, not what they wear. The model keeps each face as a likeness and then dresses and lights both people to fit the pink studio, so a sweater in a photo can turn into a T-shirt in the video. In our own renders the faces held up much better than the outfits did. If a particular outfit matters, expect to render again until it matches. Faces are the part that drifts least.

Can I post my AI APT MV on TikTok or Reels?

Yes. It is a vertical 9:16 MP4 that fits TikTok, Instagram Reels and YouTube Shorts without cropping. Two things to do when you post: make sure both people agreed to be in it, and switch on the AI-generated content label, which those platforms ask for on realistic AI video. Our audio is original to the clip, so you can keep it, or swap it for the licensed song from the platform library as described above. Tag the other person in the caption so the clip reads as a joke between the two of you rather than a mystery video.