0 to 2 seconds: the first miss
The Scooper dips the spoon into the coffee jar and tips the spoonful onto the counter beside the red mug. Behind, the Pourer picks up the steel kettle. Nobody comments.
Video trend
The duo coffee trend is a 17-second one-take with no cuts, and the joke lives in its timing: small misses, a calm face, then one big flood. Knowing the beats helps you judge your own version and choose where to put a caption or a cover.
The Scooper dips the spoon into the coffee jar and tips the spoonful onto the counter beside the red mug. Behind, the Pourer picks up the steel kettle. Nobody comments.
The Pourer fills the empty mug from the kettle with a serious face. The brown heap on the counter sits right next to it. The Scooper looks down at the heap, spoon still in hand.
The Scooper grins, laughs, scoops again and drops more coffee on the counter. The Pourer sets the kettle down and glances over. The mess is still small, which is what makes the next beat land.
The Scooper picks up the mug at the same moment the Pourer tips a small white jug, so the milk pours straight onto the bare counter where the mug stood. Within a second the whole counter is a white sheet with coffee floating in it.
The Scooper holds the red mug at chest height and gestures with the spoon as if the coffee were ready. The Pourer stands back and stares. The clip ends there, about 17 seconds in.
Oakgen's template follows the reference frame by frame with a depth motion guide and puts the original song back at the end, so your photos change who is in the kitchen, not when the milk goes.
Add a photo for each role shown below. All roles need a photo.



Upload a clear head-and-shoulders photo of the person who should spoon the coffee and lift the mug. Face forward, hair fully visible and no hat works best, because the Scooper fills the front of the frame.
Upload the Pourer's photo the same way; they stand further back, so a sharp, well-lit face matters. Choose 480p, 720p or 1080p and review the quote, which covers the full video and both preparations.
Watch that the two never swap faces, the coffee lands on the counter and the milk floods it in one pool, then download your portrait MP4 with the song.
The Scooper spoons coffee next to the mug, grins and lifts the mug at the end; the Pourer fills the empty mug from the kettle and pours the milk where the mug just stood. Each takes the face, hair and outfit from their own photo.
All 17 seconds in one handheld take: the first spoonful on the counter, the kettle pour, the second miss, the milk flood and the deadpan stare, with the original song kept.
There is no costume. The football shirts from the reference are replaced by what each person wears in their photo, and no shirt gets a crest, number or logo.
Partners, best friends, siblings, a parent and a child, or you and your dog. The reference pair are replaced by grey silhouettes in the motion guide, so neither is reused.
It is a 17-second portrait AI clip of two people at a granite kitchen counter trying to make one cup of instant coffee. The one in front spoons the coffee onto the counter beside the red mug, the one behind pours hot water into the empty mug, and when the mug is finally lifted the milk goes straight onto the counter. The Pourer never flinches and the Scooper only grins, and that is the joke. Oakgen's template puts the two people from your photos in it.
Yes. The same clip is shared as the Cofi trend (#cofitrend), the Couple Coffee Mess, the duo coffee trend, the coffee making AI video and the AI coffee spill video. They all describe two people at one counter missing the mug and flooding it with milk. Oakgen has one template for it under the name Couple Coffee Mess.
Yes. "When two legends try to make coffee" is the caption many posts use for the same kitchen skit, often with two famous faces. Here you cast two people from your own photos instead, so it becomes you and your partner, a friend or a sibling.
"Con la Misma Piedra" by Julio Iglesias, a Spanish ballad from 1982. The title means "with the same stone", from the saying about tripping over the same stone twice, which suits two people who miss the mug again and again. Your video keeps that original soundtrack, and nobody on screen sings.
No. Famous recordings are refused as reference audio, so the song is not sent to the video model. The mouths follow the motion guide, and the original soundtrack is put back at the end.
An AI-made clip with look-alikes of two football stars in their club and national-team shirts. Oakgen is not affiliated with them, their clubs or the shirt brands, and neither look-alike is kept in your video.
Yes. The Scooper and the Pourer each need a photo, because the motion guide replaces both reference people with grey silhouettes.
The Scooper stands in front on the right, spoons the coffee onto the counter, grins and lifts the red mug at the end. The Pourer stands behind on the left, pours the kettle into the empty mug and then pours the milk onto the counter. Put whoever always makes the mess in the Pourer slot.
Each person wears the outfit from their own photo. The football shirts, armbands, tattoos and watch from the reference are not kept, and no shirt gets a crest, number or logo.
A head-and-shoulders photo of one person, facing the camera, in even light, with the hair fully visible and no hat. The Pourer stands further from the camera, so a sharp face matters most there.
Yes. Either slot can be a dog, a cat or another pet; it keeps its species, coat and markings and works the counter with its paws. If you want two pets in the original golden retriever kitchen, use the Dog Coffee Fail template instead.
Both are versions of the same coffee-making joke. The Dog Coffee Fail is the pet version on a black counter, ending with one dog covering its eyes. The Couple Coffee Mess is the two-person version on a granite counter, with the coffee spooned beside the mug and the milk poured where the mug stood.
Your own, or photos of people who agreed to appear. Do not upload photos of footballers, celebrities or anyone who has not consented, and label AI videos clearly when you share them.
About 17 seconds in portrait 9:16, one continuous handheld take from the first spoonful to the deadpan stare.
No. It is the AI reference clip the trend comes from, used as the motion and timing reference, not an Oakgen result. Its two people are never reused.
3,596 credits per 720p video. The quote includes both photo preparations and the complete video. Check it before generating.
Plan a story, keep your character recognizable, and make the finished clip your own.
Two versions of the viral coffee-making AI joke side by side: who they suit, what each photo controls, the kitchen, the ending and which template to pick.
Make the Couple Coffee Mess AI video, the Cofi trend, step by step: pick the Scooper and the Pourer photos, read the quote, check the milk flood and post it.
What the Cofi trend is, why it also goes by Couple Coffee Mess, duo coffee and "When two legends try to make coffee", what the song is and why the joke works.
Upload 2 photos. Everything else is set.