Baby Lip Sync AI is an AI video that puts you, and up to two friends, into the pass-the-camera Baby lip sync trend: you sing Justin Bieber's Baby into a handheld phone, then hand the camera on. videosi makes it in the browser as a 12-second MP4 with the trend sound on it. This guide goes deeper than the trend page: where the format comes from, how it spread, and what to expect from your own clip.
Key takeaways
- The trend is a pass-the-camera lip sync to Baby by Justin Bieber featuring Ludacris (2010), first seen on TikTok in July 2026.
- The earliest clips we found are dated 28 July and 14 August 2026; a UFC 333 backstage version on 24 October 2026 gave it the "wait for Dana's part" twist.
- Up to three people come from photos, in order; the rest of the crowd is made up, and everyone is an adult. The price is the same with 1, 2 or 3 photos.
- The clip has the Baby sound on it, but mouths do not match the lyrics exactly.
- A standard 12-second clip is 10 credits; packs start at $6.99 for 10 credits, with no free trial.
In this guide
- What is the Baby lip sync trend?
- Which song is it, and is it the original Baby?
- Where did it start, and how did it spread?
- What does "wait for Dana's part" mean?
- Who is in the original clips, and are they real?
- How do I make the Baby lip sync AI video from photos?
- How many people can be in it, and who gets which turn?
- Why are only adults accepted?
- Does it really lip-sync the lyrics?
- How much does it cost, and is there a free version?
- Why doesn't my clip look like me, and how do I fix it?
- Is the trend still going, and where should I post the result?
What is the Baby lip sync trend?
The Baby lip sync trend is a pass-the-camera video. One person holds a phone at arm's length and lip-syncs the chorus of Baby straight into the lens, then hands the phone on, and the next person picks up the line. The clip keeps going through a whole group, and the joke is how many people in a row know every word. It looks effortless, which is why friend groups, gyms and offices all copied it. The format needs no editing at all, only a crowd that is willing to sing badly with confidence.
Which song is it, and is it the original Baby?
It is Baby by Justin Bieber featuring Ludacris, released on 18 January 2010 according to its Wikipedia entry. The clips use the original recording, so the chorus is instantly recognisable even when the singers are not in tune. Some posts show the track as Baby in TikTok's sound listing and others use a re-uploaded original sound, so the exact upload differs from clip to clip. The videosi clip carries the sound taken from the trend's original video.
Where did it start, and how did it spread?
The earliest pass-the-camera Baby clip we found was posted on 28 July 2026 by a warehouse crew. A bigger version, a large group in an old hangar, followed on 14 August 2026. The biggest push came on 24 October 2026, when TNT Fight Sports posted a UFC 333 backstage clip in which fighters pass the phone around a packed room. Those dates are the earliest we could confirm, not proof that any one post invented the format, so treat the origin as unconfirmed.
What does "wait for Dana's part" mean?
It is the caption on the UFC 333 clip. The fighters sing in turn and the last turn goes to UFC president Dana White, who is the least likely person in the room to be singing Baby. That became the twist other groups copied: save the most unlikely person for last, then cut when the camera reaches them. It is a joke about expectations, not a rule of the format, and you can end your own round on anyone.
Who is in the original clips, and are they real?
The warehouse, hangar and UFC clips are real phone videos of real adults, not AI. The AI version is a separate thing: videosi swaps the people for you and your friends and invents the rest of the crowd. It does not use the faces from those clips, and it does not accept photos of celebrities, athletes or other public figures. Everyone in your clip is either from your own photos or a made-up adult.
How do I make the Baby lip sync AI video from photos?
Open Baby Lip Sync AI on videosi, add one front-facing photo of yourself, optionally add a friend's photo for the second turn and another for the third, pick 9:16, 1:1 or 16:9 and press Generate. The result is a 12-second MP4 with the Baby sound already on it and no watermark. You write no prompt, because the warehouse, the hand-offs and the crowd are built in.
How many people can be in it, and who gets which turn?
Up to three people from photos, in the order you add them. Photo 1 sings first, photo 2 takes the camera at about 4 seconds and photo 3 at about 8 seconds. Any turn without a photo is sung by a made-up adult, and so is the crowd behind them. Adding friends costs nothing extra. The original videos run 35 seconds or more with eight or more singers, which does not fit into one 12-second render.
Why are only adults accepted?
The clip is built for adults only. The prompt requires everyone in it to be clearly over 18, and videosi does not accept photos of anyone under 18. The trend itself has plenty of group videos at gyms, workplaces and sports teams, and the AI version follows that adult setting. Only add friends who agreed to appear, and use a separate photo for each person rather than a group shot.
Does it really lip-sync the lyrics?
Not word for word. The Baby sound is on the clip and every mouth moves as if singing, but the model does not match mouth shapes to the lyrics. At 12 seconds the clip sells the hand-offs, the faces and the pass-the-camera rhythm, so it reads as a performance. If you want exact lip-sync, record the real thing with your friends; that is also how the original trend works.
How much does it cost, and is there a free version?
There is no free generation and no free trial. A standard 12-second video (Standard model, 720p) is 10 credits, and better models and higher resolutions cost more. Credit packs start at $6.99 for 10 credits and never expire. A failed, timed-out or blocked render returns its credits automatically, and the price is the same with 1, 2 or 3 photos.
Why doesn't my clip look like me, and how do I fix it?
The usual cause is the photo. Sunglasses, a cap, a side angle or a dark, grainy shot all hurt likeness. Use one front-facing photo with open eyes, head and shoulders, in even light. Likeness drifts most when the camera swings from one singer to the next, around seconds 4 and 8, so put the person who matters most in photo 1, who gets the longest and clearest shot.
Is the trend still going, and where should I post the result?
As of 1 November 2026 the Baby lip sync trend is still spreading on TikTok and Instagram, mostly as friend-group and workplace rounds. Post the MP4 as a vertical video, label it as AI-generated, and keep the trend sound on it. If the platform mutes the track, add Baby again from the app's own sound library. videosi is independent and not affiliated with Justin Bieber, UFC, TikTok or Instagram.
Ready to make your own?
Pick your best front-facing photo, add two friends if you like, and open Baby Lip Sync AI. For other songs and group formats, the Tiny Desk AI page works the same way with a band and a sax solo.
