Quick answer
Upload one photo to Starrd's I Wanna Love You template and it turns you, or your pet, into the viral selfie lip-sync. You hold the camera yourself and sing to the lens up close, then lift it higher and higher as you bend all the way back and belt "already know I wanna love you" straight up into it, then swing back up and wave. It follows the original clip's moves, timing and mouth shapes line by line, keeps your eyes on the lens the whole time, and plays the song over the top. Your own face and outfit, in the setting from your photo. 15 seconds, vertical 9:16, 100 credits (about $5).
One photo in, and out comes you, or your pet, fully committed to a selfie lip-sync: a close-up serenade, a full backbend belt, and a wave. Here's one made from a single photo of a husky:
I Wanna Love You
One photo, you or your pet. A selfie lip-sync that ends in a full backbend belt. 15 seconds, vertical.
What You Get
- Selfie mode. The whole video is your own front camera, held in your own hand or paw. No phone in the shot, no second person filming.
- The serenade. Up close to the lens, head swaying, eyebrows up, singing every word.
- The backbend. The camera rises over your head as you bend all the way back and belt the chorus straight up into it.
- The sly look. A smug, smouldering beat while the song goes quiet.
- The wave. Back up to the front, waving along to the last line.
- Eyes on the lens. The whole way through, like you're singing to whoever is watching.
- Your own setting. The room from your photo, not a borrowed one.
- 15 seconds, vertical 9:16, one continuous take with the song, ready for TikTok, Reels and Shorts.
What the Trend Actually Is
The original is a front-camera clip of a man in a bowl-cut wig and a fake moustache lip-syncing Akon's "I Wanna Love You" outside at night. He starts cool and close to the lens, then keeps lifting the phone higher while he leans back, and back, until he is flat on the ground belting the chorus up at it, before springing back up for the last line.
Why it suits AI: the moves are simple, but the commitment and the timing are everything. One photo gets you the full performance on the beat, and it works on your dog, which is where it gets funny.
The Beats
| Time | What happens |
|---|---|
| 0–4s | Steps in already singing, close to the lens |
| 4–5.5s | "I know you see me", head swaying, eyebrows up |
| 5.5–7.5s | Lifts the camera over their head and starts leaning back |
| 7.5–11s | Bent all the way back, belting "already know I wanna love you" up into the lens |
| 11–12.5s | A sly, smouldering look while the song goes quiet |
| 12.5–15s | Swings back up to the front and waves along to the last line |
People or Pets
The template reads your photo. A person holds the camera at arm's length and does the moves as written. A dog, cat or any other animal films with its own front paw, sings with its own mouth (real teeth and tongue, no human lips) and gets extra expression: big adoring eyes on the serenade, a wide-open belt, a smug look on the quiet beat and a cheeky grin on the wave.
Use one clear, well-lit photo where the face is sharp and the background is somewhere you'd actually film a selfie. The background becomes your set.
The Fastest Way
I Wanna Love You
One photo, the full 15-second selfie lip-sync with the song. People and pets.
One photo, 100 credits (about $5), no prompt. It renders on Seedance 2.5 at 480p, vertical 9:16.
Build It Yourself
This one is driven by the original clip, not a written story. The moves come from the clip; the work is in the lip sync.
1. Make a character sheet. One big selfie panel at the top, set in the photo's own background, with the arm reaching toward the lens from the right edge. Below it, a 360° turnaround of the subject on white: front, three-quarter, side and back.
A photorealistic vertical 9:16 character sheet of the person from the photo. Top two-thirds: their own front-camera selfie in the same room as their photo, their left arm reaching toward the lens from the right edge of the frame, eyes looking straight into the lens, mouth open mid-song. Nothing holding the camera is visible. Bottom row: a 360° turnaround on a plain white background, four full-body views, front, three-quarter, side and back. Same face, hair and outfit in every panel.
2. Drive it with the original clip, song included. Use the clip as the motion reference with its audio still in it, so the model hears what it is syncing to.
3. Write one beat-by-beat list. Each line gets the time, the move, the face and the exact lyric together. Write "mouth closed" on the beats with no vocal, and say the mouth keeps singing through every syllable with no pause mid-line.
4. Never write the word "phone". Say "selfie camera". Every time you mention a phone, the model draws one in the shot.
Common Mistakes
A phone in the frame. The camera is the phone. If you can see one, the prompt mentioned it.
Switching hands. Pin the camera to one arm and one edge of the frame, or it swaps sides halfway through.
Eyes drifting. "Eyes half-closed on the long notes" turns into closed eyes. Say the eyes stay open and on the lens in every beat.
An upside-down face. Show the backbend in the clip, not in the sheet. A sheet panel of the subject upside down gets copied as a flipped shot.
FAQ
How many photos? One.
People or pets? Both.
How long is it? 15 seconds, vertical 9:16, one continuous take.
What does it cost? 100 credits, about $5.
Related Reading
- How to Make the 'I'm Going To Kill You' Mirror AI Video: another one-take meme, starring you or your pet
- How to Make the 'Me At Work' Keyboard Smash AI Video: a deadpan meme, starring you or your pet
- How to Make the Burning Bridges AI Video: another lip-sync performance to a real song
- Viral AI Video Trends (2026): The Monthly Roundup: every trend worth making this month
