Quick answer
Starrd's Nazeer Bakra Goat template puts you in the viral clip from TikTok creator @nazeer.bakra.offi. Upload one photo. You sit on a cement rooftop wall above a bright green rice field and let out wordless, goat-like cries straight into a phone with your tongue out and your eyes screwed shut. You turn away, howl, laugh silently, give a smug little smile, howl once more, then throw one arm in the air and grin as the camera pulls back. It's one continuous take, the motion is copied from the original clip as a depth map so every cry lands on the original's timing, and the original sound is on it. Your face, hair, glasses and outfit colours come from your photo, while the clothes take a looser shape from the original. Pets work too, sitting upright on the ledge. 12 seconds, vertical 9:16, Seedance 2.5, 480p, 150 credits (about $7.50).
Tongue out, eyes screwed shut, and a noise no person should be making, delivered with total commitment. One photo puts you on the wall in the viral Nazeer Bakra goat clip. Here it is:
Nazeer Bakra Goat
One photo. You bleat into the phone on a rooftop above a rice field, then throw one arm in the air and grin. 12 seconds.
What You Get
- One unbroken take. No cuts: a handheld phone stays close to your face the whole way, then backs off at the end.
- Every cry on time. Your mouth opens on each cry and closes where the original's does, because the sound is part of what drives the render.
- The tongue. It comes out over your lower lip three times, on the original's timing.
- The silent laugh. About two seconds of nothing, mouth hanging open, one hand swinging low.
- The payoff. One last long cry, an arm thrown straight up, a huge grin, and the camera pulling back to show you perched on the wall.
- People or pets. A dog or cat sits upright on the ledge and howls with its own mouth.
- 12 seconds, vertical 9:16, with the original sound, ready for TikTok, Reels and Shorts.
What the Video Actually Is
The clip is from TikTok creator @nazeer.bakra.offi, posted on September 18, 2026 (10.1M+ plays, 233K+ likes and 23K+ saves so far). A man in a mint-green kurta sits on a low cement wall on a rooftop, with a bright green rice field stretching out behind him, and lets out a run of wordless cries straight into a phone. The sound has no words at all, just short yelps and long, wavering bleats with hard silence between them. "Bakra" means goat in Hindi and Urdu.
Why it suits AI: the whole joke is the commitment, and that gets funnier on a face people know. You don't need the rooftop, the rice field or the nerve to film it. One photo covers all three.
The 12 Seconds, Beat by Beat
| Time | What happens |
|---|---|
| 0:00 | You look into the lens, mouth half open |
| 0.2s | A short yelp, and the tongue starts to come out |
| 0.6s | Head tips back, eyes screw shut, tongue out over the lower lip, the first long cry |
| 1.5s | A long, wavering cry through a half-open mouth |
| 2.4s | Lips shut, and at 2.5s the head turns sharply away |
| 2.9s | The head snaps back to the lens, the tongue flicks out, a short cry |
| 4.1s | Chin up, the mouth at its widest, the longest cry yet |
| 5.1s | Lips shut in a scrunched grimace |
| 5.5s | One more short cry |
| 5.9s | Silence: mouth hanging open in a breathless laugh, one hand swinging low from side to side |
| 8.2s | Two quick cries, tongue out on the second |
| 9.1s | Lips closed in a smug little smile, eyebrows raised |
| 9.6s | The last long cry, as one arm swings up from 9.9s and stays there |
| 10.7s | The cry ends and the mouth settles into a huge grin |
| 11.1s | The camera backs away to show you perched on the wall, one knee up, the other leg hanging down |
Your Look, Their Moves
- Your face and hair. The original's head and beard stay out of it, so you keep your own hairstyle and facial hair.
- Your details. Sunglasses, a scarf and tattoos all carried over in our test.
- Your colours, a looser cut. The clothes keep your outfit's colour, but can take a looser shape from the original's tunic and trousers. Our test photo's pink suit came out as a dusty pink tunic, which we think suits the joke.
Pick a clear, well-lit photo where your face is easy to see, because most of this clip is a close-up of your face. A full-length photo helps too: the last second pulls back to show your whole body on the wall.
The Fastest Way
Nazeer Bakra Goat
One photo, one take: bleat into the phone, throw an arm in the air and grin. 12 seconds.
One photo, 150 credits (about $7.50), no prompt. It renders on Seedance 2.5 at 480p, vertical 9:16, 12 seconds, with the original sound on top.
Build It Yourself
1. Turn the clip into a depth map, and keep the sound in it. Run the clip through a depth-estimation model (Depth Anything V2 works). The depth map keeps the head moves, the raised arm and the camera pull-back, and drops the face, the beard, the clothes and the colour. Put the clip's sound in the depth video, so the model has something for the mouth to follow.
2. Make one character sheet, not a storyboard. This clip is one take, so you don't need a sheet of action stills. Generate a single reference image from your photo: the empty rooftop, then a front, three-quarter, side and back view of you. In our storyboard test, the render jumped straight to a drawn panel at 11.75 seconds. The character sheet version ran clean, with no jump.
A photorealistic vertical 9:16 character reference sheet laid out as two bands separated by thin black lines, with no numbers, labels or text. Everything is set on the flat roof of a farmhouse on a bright overcast day: a low rough cement parapet wall with a wide ledge and a vast field of vivid green rice behind it. Top band: the rooftop with nobody on it. Bottom band: four equal panels of the person from the reference image, full figure from head to feet, from the front, three-quarter, side and back, relaxed and still, mouth closed, arms hanging at the sides.
3. Drive it with the depth map. Feed the sheet, your photo and the depth video to a model that copies motion from a reference video (we recommend Seedance 2.5). Then write what it can't read from a depth map:
- The cries: a timed list of when the mouth is wide open, when the lips are shut, and the silent stretch from 5.9 to 8.2 seconds.
- The tongue: the three moments it comes out, because a depth map doesn't show one.
- The ending: no more cries after 10.7 seconds, just a still, huge grin.
Lay the sound back on top at the end: Seedance can drop or change audio on its own, so mux the track yourself.
Common Mistakes
Driving with the raw clip. The original performer's face, beard and clothes come straight through. Use the depth map.
Using a storyboard on a one-take clip. Drawn panels can come back as a sudden jump mid-shot. A character sheet gives the model the look without giving it another shot to cut to.
Trusting a transcript. Speech-to-text returned noise in four languages, because there are no words. Time the cries from the waveform and the mouth, frame by frame.
A silent depth map. Without the sound in it, the mouth has no cue and can drift away from the cries.
Filling the silence. Without a timed window, the model keeps the character crying through the quiet stretch in the middle. Say plainly that nothing happens from 5.9 to 8.2 seconds except a hanging mouth and a swinging hand.
Writing "goat" in the prompt. We left the word out and described raw, wordless cries instead, so the model has no reason to add an animal to the rooftop.
Related Reading
- How to Make the 'Big River Flows East' (大河向东流) Monkey Meme Starring You or Your Pet: another clip where the joke is a wild animal reaction, from one photo
- How to Make the 'Bombaclat Wallahi' AI Video: The Food Truck Meme Starring You: one photo in a viral food truck argument, lip-synced to the original
- How to Make the Stadium Calm Down AI Video: another one-take meme from one photo, this time in a roaring stadium
- How to Make the Lil Yachty Walkout AI Video: the hardest walkout ever, in one take
- Viral AI Video Trends 2026: every trend we've turned into a template this year
