Funny AI videos from your own clips: a giant lion handshake, a scrunch face, a big head on a tiny body and a motor-mouth

Four funny AI videos made from plain clips of yourself, each with one gag added.

9 generations$2.89 total

Disclaimer: This is an independent creative test made with AI, for demonstration only. It isn't an ad for any brand shown or named, and isn't affiliated with, sponsored or endorsed by them. All trademarks belong to their owners

This is the video I posted on Instagram, here's the link: https://www.instagram.com/reel/Df5Qj4EMHgN/

In this post you'll learn how to:

  • shake hands with a giant lion that actually sits in your shot instead of looking pasted on
  • scrunch your whole face up to a ridiculous degree as your head leans into the lens
  • put a big head on a tiny body as one believable little figure, not a cutout
  • turn your face into a giant mouth that talks a mile a minute while your body stays alive

I keep taking totally normal clips of myself standing around outside and adding one dumb thing to make people stop scrolling. A giant lion, a scrunched-up face, a big head on a tiny body, and a mouth that won't stop talking. Every one started as a plain 3 to 4 second phone clip. None of it is a filter you download, it's one AI edit added to my real footage, so my background, my clothes and my movement all stay exactly as I shot them.

Here's my exact process.

1. Shake hands with a giant lion

I filmed myself holding one hand out to the side, nothing there. Then I grabbed the frame where my hand is open and built a still that drops a giant lion in front of me, lowering its paw into my hand.

A man in a plaza holds his hand out and a giant lion lowers its paw into his open hand., made with GPT Image 2.6 Flare

Exact prompt

Photoreal edit of the reference photo. Keep the reference photo exactly as it is: same man, his black t-shirt, dark pants, sneakers, his profile stance with his right arm extended forward and his hand open, the same palm plaza, the daylight and framing. Preserve his face exactly, no aging. Add ONE thing: a giant lion placing its paw in his open hand. THE VIBE: A funny, epic visual hook: a giant lion, far bigger than a normal lion, faces him and reaches out one huge paw to rest it gently in his open hand, like the two of them are shaking hands. THE SUBJECT: Keep the man exactly as in the reference, his right arm extended forward and his hand open. In front of him, in the direction his extended hand points, a giant photorealistic lion faces him: much larger than a real lion, its head and mane towering above the man's height. The lion sits or crouches upright and extends one enormous front paw forward, resting that paw in and against the man's open hand, like a calm paw-shake. Detailed golden fur, full mane, majestic and gentle. The paw meets his hand at the same height his hand is held. THE FACE: Preserve his face exactly from the reference, no aging. THE SETTING: The same palm plaza, buildings, people and daylight as the reference. THE LIGHT: The same bright daylight; the giant lion lit to match and casting a soft shadow on the pavement. THE KEY DETAILS: A giant lion facing him, one huge paw placed in the man's open hand, a paw-shake between them. THE MOOD: Funny, epic, wholesome. THE COMPOSITION: Keep the man where he is in profile; the giant lion faces him from the direction his hand points, their paw and hand meeting between them. THE CAMERA: Same phone capture and framing as the reference. No text, no logos.

The thing that sells it: build the lion into the same light as your clip and give it a real shadow on the ground. If you just paste a lion in, it looks like a sticker. I say the lion sits facing me, lit by the same daylight, with a shadow under it. Then I run the clip and the still together in edit mode so my real arm stays and the lion holds the handshake.

Kling 4.0 $0.60

Exact prompt

Create a new video where a giant lion sits in the plaza facing the man and rests its huge paw in his open hand, like a paw-shake, while the man holds his hand out. [SCENE 1] Keep everything from @video_1 exactly: the man, his face and black outfit (no aging), the palm plaza, the daylight, the camera and the timing of his motion. The giant lion from @image_1 sits facing him from the very first frame. As the man holds his hand out, the lion gently lowers and rests its enormous paw into the man's open hand in a calm paw-shake, and the paw stays resting in his hand. The lion breathes and its mane shifts slightly in the breeze. Keep his face exactly as in @video_1, no aging. @image_1 is the look to match: the giant lion and the paw-shake.

2. Put a big head on a small body

The big head small body gag is an easy one. I filmed myself standing normally, pulled a full-body frame, and built a still with my head blown up about three times too big on a tiny little body.

A man standing in a plaza with a giant head on a tiny body as one figure., made with GPT Image 2.6 Flare

Exact prompt

Photoreal edit. @image_1 is the base: keep the same plaza with the restaurant, the cars, the daylight and framing. @image_2 is his face only: match his face, beard and glasses exactly, no aging, and take nothing else from it. Change one thing: give him a giant head on a tiny body, built as ONE seamless, believable figure. THE VIBE: The big-head-tiny-body look, but it must read as one real cohesive little character, not a head pasted onto a body. THE SUBJECT: His head is enormous, about three to four times too big, sitting directly and naturally on a tiny shrunken body. The head joins the body through a short natural neck and little shoulders, with the same skin tone and the same lighting across head and body, no hard cut-out edges, no gap, nothing floating. The tiny body wears his black t-shirt, dark pants and sneakers, with short little arms and legs. The whole thing is one believable caricature figure standing on the plaza ground, grounded by a soft contact shadow under its feet. The giant head keeps his real face, beard and glasses. THE FACE: Match his identity exactly from @image_2 (face, beard, glasses), no aging. THE SETTING: The same plaza, restaurant, cars and daylight as the reference. THE LIGHT: The same daylight and sun direction across the whole figure, so the head and the tiny body share one consistent light, with a real shadow on the ground. THE KEY DETAILS: A giant head seamlessly attached to a tiny body as one cohesive figure, matched lighting, grounded shadow, not a pasted cut-out. THE MOOD: Funny, absurd. THE COMPOSITION: Centered, the giant head on top and the tiny body below, standing on the ground. THE CAMERA: Same phone capture and framing as the reference. No text, no logos.

The trick is making it one believable figure instead of a head floating over a body. I ask for the head joined to the body with matched lighting and a shadow on the ground, so it reads as a real little character. In the video the giant head bobbles on the tiny body as I move.

Kling 4.0 $0.60

Exact prompt

Create a new video where the man stands in the plaza with a giant head on a tiny body, the head bobbling slightly as he moves. [SCENE 1] Keep everything from @video_1 exactly: the plaza, the restaurant, the cars, the daylight, the camera and the timing of his motion. The man has a giant head on a tiny body exactly like @image_1, one cohesive figure. As he moves with his real motion, the giant head bobbles and tilts slightly on the tiny body and the little arms move naturally, the figure staying grounded with its shadow on the ground. Keep his face exactly: @image_2 is his face, match it, no aging, take nothing else from it. @image_1 is the look to match: giant head, tiny body, one believable figure.

3. Scrunch your whole face up

For this one I leaned toward the camera. I pulled the frame mid-lean and built a still where my neck stretches my head up to the lens and my face scrunches up hard.

A man on a beach with his head stretched to the camera and his face scrunched up tight., made with GPT Image 2.6 Flare

Exact prompt

Photoreal edit. @image_1 is the base: keep the same man, his dark shirt, jeans, the same beach and sky and light, the framing. @image_2 is his face only: match his face, beard and glasses exactly, no aging, and take nothing else from it. Change two things: extend his neck forward toward the camera, and scrunch his whole face up as extremely as possible. THE VIBE: A funny visual hook: his neck stretches his head up close to the camera and his entire face is scrunched up to the absolute maximum, compressed into a tight ball of deep wrinkles, way beyond a normal scrunch. THE SUBJECT: Keep his body on the beach. His neck stretches and extends forward toward the camera, bringing his head large and close to the lens. His face is scrunched up to the extreme: eyes squeezed so tightly shut they almost vanish into deep creases, nose wrinkled up hard with heavy folds across the bridge, cheeks bunched up high and round, eyebrows crushed down and inward, the forehead and the whole face compressed inward with deep wrinkles everywhere, lips mashed together and puckered hard forward. The whole face looks squished and compressed into a silly tight knot, as extreme as someone sucking a lemon as hard as they possibly can. Keep his beard and glasses. THE FACE: Match his identity exactly from @image_2 (face, beard, glasses), no aging, just scrunched up to the extreme. THE SETTING: The same beach, sky and light as @image_1. THE LIGHT: The same daylight; his close-up face lit to match. THE KEY DETAILS: Neck extended forward, head large and close to the camera, face scrunched to the absolute maximum with eyes buried shut, deep wrinkles everywhere and lips mashed forward. THE MOOD: Funny, goofy, extreme. THE COMPOSITION: His stretched neck leads his head toward the camera, his heavily scrunched face filling much of the frame. THE CAMERA: Same phone capture; his face is now close because of the neck extension. No text, no logos.

Push this way further than feels reasonable. A gentle scrunch reads as nothing, so I ask for the whole face compressed: eyes squeezed shut, nose wrinkled, deep creases, lips mashed forward. In the video it builds with my real lean, so the scrunch lands right on the motion.

Kling 4.0 $0.88

Exact prompt

Create a new video where the man leans toward the camera, his neck stretching his head up to the lens as his face scrunches up extremely tight. [SCENE 1] Keep everything from @video_1 exactly: the man, his clothes, the beach, the light, the camera and the timing of his motion. As he leans forward with his real motion, his neck stretches and extends his head forward toward the camera and his face scrunches up to the extreme exactly like @image_1: eyes squeezed shut, nose wrinkled, cheeks bunched, lips mashed and puckered forward, deep wrinkles. By the peak his scrunched face is large and close to the lens, then it eases back as he leans back. Keep his identity exactly: @image_2 is his face, match his face, beard and glasses, no aging, take nothing else from it. @image_1 is the look to match.

4. Turn your face into a giant talking mouth

Last one. I pulled a frame of me sitting, then built a still where my face is almost entirely one giant open mouth, everything else shrunk to the edges.

A man sitting on a ledge whose face is almost entirely one giant open mouth., made with GPT Image 2.6 Flare

Exact prompt

Photoreal edit. @image_1 is the base: keep the same man, his shirt, jeans, the stone ledge and the hedge behind him, the light and framing. @image_2 is his face only: match his beard, hair and skin tone exactly, no aging, and take nothing else from it. Change two things: extend his neck forward toward the camera, and turn his face into almost entirely one giant open mouth. THE VIBE: A funny visual hook: his face is basically just a huge gaping open mouth. The open mouth is so big it takes up nearly his whole face, with everything else shrunk to the edges, exactly like a face that is mostly mouth. THE SUBJECT: Keep his body sitting on the ledge with his arms spread. His neck stretches and extends forward toward the camera, bringing his head large and close to the lens. An enormous wide-open mouth fills almost the entire face, from near the top down to the chin, showing big upper and lower teeth and the tongue inside. His eyes are shrunk into two small slits pushed right up to the very top edge of the face, just under his hairline, with almost no forehead left. His beard frames the outer edges and jaw around the giant mouth, and his hair sits above. The mouth clearly IS most of the face. THE FACE: Keep it recognizably him through his beard, hair and skin tone from @image_2, no aging, but with the giant mouth taking over the whole face and the eyes reduced to tiny slits at the top. THE SETTING: The same ledge, hedge and daylight as @image_1. THE LIGHT: The same daylight; his close-up face lit to match. THE KEY DETAILS: The face is almost entirely one giant gaping open mouth with big teeth and tongue, eyes reduced to small slits at the very top, neck extended forward to the lens. THE MOOD: Funny, wild, over-the-top. THE COMPOSITION: His stretched neck leads his head to the camera, the giant mouth dominating almost the whole face. THE CAMERA: Same phone capture; his face is now close because of the neck extension. No text, no logos.

Two things make this work. One, let the mouth take over most of the face instead of just being a big mouth on a normal face. Two, in the video make it actually talk, nonstop and fast, jaw flapping the whole time, because the whole joke is the mouth. And keep your body moving naturally while it yaps, or the clip feels frozen.

Kling 4.0 $0.61

Exact prompt

Create a new video where the man sits on the ledge, his face a giant gaping mouth talking a mile a minute, and his whole body moves and gestures naturally and lively as he talks. [SCENE 1] Keep from @video_1: the same man, the stone ledge, the hedge, the daylight, the camera and his overall seated position. His face is almost entirely one giant open mouth exactly like @image_1, and the mouth is TALKING nonstop and extremely fast, the jaw flapping wide open and shut over and over. As he talks, his whole body moves naturally and lively: he leans forward and back, shifts his weight on the ledge, rolls and shrugs his shoulders, bobs his head and gestures with his hands and arms, animated like an excited fast-talker, loose and never stiff or frozen. Keep his identity: @image_2 is his face, match his beard, hair and skin tone, no aging, take nothing else from it. @image_1 is the look to match.

Why it works

Every one of these is the same move: take a plain clip of yourself and add exactly one impossible thing on top, in edit mode so the model keeps your real footage. My camera, my timing, my clothes and my background all stay, which is why a giant lion or a huge mouth sits in the shot like it was really there. I build the look as a still first, cheap, get it right, then spend on the video once the still is good.

You can do this with any clip of yourself. Film 3 to 5 seconds of one simple action, pick the one gag you want to add, build the still, then animate it. Keep each clip short, test the still before you pay for the video, and stitch them together at the end.

Your turn: grab one 3 second clip of yourself and add one dumb thing to it. The big head on a tiny body is the most forgiving place to start.

Emero runs all of this at cost and remembers your face between clips, so the next gag is faster to make.

Here's the full cut one more time:

Total for this workflow$2.89

Get one of these every day

I test new workflows daily and send them with the exact prompts, the models and the costs

6,526 generations tested so far

More workflows

See all