How to Make the APT. AI Music Video From Two Photos

A bright pink studio, a red drum kit, one person behind the drums and one with the mic. Two-person APT. swaps of that set spread across social feeds, including clips featuring Sam Altman and Dario Amodei. Here is how to put you and a friend on the same stage from two photos, without a prompt, a filter or a green screen.

What is the APT. AI trend?

APT. by ROSÉ and Bruno Mars was released on 18 October 2024, and its official video plays out in a pink studio with a drum kit. The AI trend swaps two new people into that studio: one plays the drums, the other sings, and both dance through the same shots. You'll also see it searched as APT MV AI or APT AI MV. The APT. Pink Studio template puts the people from your two photos in those roles, with the song. It is an independent take on the trend; the viral founder clips were not made by OpenAI or Anthropic.

How to make it

1. Open the APT. Pink Studio template.
2. Add photo 1: the drummer, who plays the kit and gets the sunglasses gag.
3. Add photo 2: the singer, who holds the mic, kneels on the pink floor and dances beside the drummer.
4. Pick a version and generate. Premium makes the full video of about 26 seconds; Balanced makes the first 14 seconds as a highlight.
5. Download the 16:9 MP4 and post it to TikTok, Instagram Reels or YouTube Shorts.

What happens in the video

• The opening: the singer kneels on the pink floor in front of the bass drum, sings into the mic and reaches toward the lens, while the drummer plays behind the kit.
• At the kit: close two-shots of the pair, the sunglasses gag, and the singer hugging the drummer from behind.
• The dance: both step out in front of the kit and dance side by side, full body, on the pink floor. The Balanced highlight ends here, at about 14 seconds.
• On the floor (full version): the two sit cross-legged facing each other and play a clapping game, then lean into the lens together.
• The ending (full version): close-up laughs, and the two stand side by side, turn to each other and laugh at about 26 seconds.

Short or full?

There is one template with two lengths. Balanced renders the first 14 seconds: the opening, the drum-kit shots and the dance. Premium renders the whole performance of about 26 seconds, including the floor scenes and the laughing ending. The roles stay fixed in both: photo 1 is always the drummer, photo 2 always the singer.

Photos that work best

• One person per photo, clear and front-facing, evenly lit, nothing covering the face.
• Pick your roles. Swap the photos before you generate if you'd rather sing than drum.
• Everyday photos. JPG, PNG, WebP or HEIC under 20 MB each.
• Ask first. Use photos of yourself and friends who agreed to it. Photos of politicians, celebrities or other public figures, and film or music-video stills, won't pass review.

What does it cost?

Premium (the full ~26-second video) is 45 credits at 480p, 100 at 720p and 180 at 1080p. Balanced (the 14-second highlight) is 15 credits at 480p and 25 at 720p. Without an account, one video starts at $7.99 at checkout; credit packs start at $9.90 for 20 credits, with no subscription, and credits never expire. No watermark, and failed renders are refunded automatically.

More on APT.: APT. CapCut template vs AI · can you make it for free?

Make your APT. video →