The AI rich video trend is a 14-second money flex in which a person hugs a long slab of cash, sweeps it across the lens, and ends up behind a table buried in bundles, fanning notes at the camera. None of it is real: on videosi you make it from one photo of yourself, press Generate and download a silent MP4. The earliest version we found was posted on TikTok on 14 October 2026, and the original creator is unconfirmed.
Key takeaways
- The table fills with cash during the sweep across the lens, which hides the change, and a low camera makes the money look huge.
- Fanned notes and hands are where AI detail fails first, so expect to retry sometimes.
- A standard 14-second video (Standard model, 720p) costs 12 credits, with no free generation, and failed renders return credits.
- Label the post as AI-generated and never use the clip to claim real wealth or sell an investment.
In this guide
- Why is the camera so low, right at table height?
- Why does the money sometimes seem to melt or merge?
- How does the table fill with cash in a single shot?
- How do I choose the right photo?
- How do I choose between the two rooms?
- How is this different from the CapCut template version?
- What does the video model actually do with my photo?
- Do I need to label the video as AI when I post?
- Can I use this clip to look rich or to sell something?
- What does it cost, and what happens if it fails?
Why is the camera so low, right at table height?
Because the low angle does three jobs at once. First, it makes the cash look huge: a slab of bundles held close to a lens placed just above the table surface fills the frame and feels like a wall of money. Second, it hides the thing that makes the reveal work. The table sits at the bottom edge of the picture, so you never see the surface clearly until the arm has swept across the lens and the bundles are already there.
Third, it looks like a phone propped on a desk, which is how real flex clips are filmed. That ordinary, slightly amateur framing is what makes a generated video believable for a few seconds. A polished cinematic angle would look staged and the joke would land worse.
Why does the money sometimes seem to melt or merge?
Banknote bundles are one of the hardest things for a video model to draw. Each bundle has a rigid shape, a paper strap and edges that must stay straight while hands move around them. When the model is unsure, it blends neighbouring objects together, so a pile can slowly turn into a green smear or a bundle can sink into a palm.
undefined
The built-in recipe fights this in two ways. It tells the model that the cash behaves like stiff paper, with no melting, merging or notes passing through hands, and it uses a motion reference so the model copies a real timing instead of inventing one. It cannot remove the problem entirely. If a render has a fan that looks wrong, that is a normal failure mode of current AI video, and a second try often looks cleaner.
How does the table fill with cash in a single shot?
This is the trick the whole trend hangs on, and it is a clever one. For about a second, between roughly the sixth and seventh second, the person swings the ribbon of money across the lens. For a few frames the picture is only an arm and cash. When the arm clears, the table in front of them is covered edge to edge with bundles, and their hands are empty.
There is no cut, but the sweep acts like one. The covered lens gives the model room to change the scene without the viewer seeing the change. It is the same idea as a hidden cut in a filmed one-take, here used to make hundreds of bundles appear.
How do I choose the right photo?
Your photo is the only source for your face, hair, build and clothes, so it carries more weight here than in most trends. Choose one front-facing picture in soft, even light, with your head, shoulders and ideally both arms in the frame. The video is almost entirely hands and money, so a waist-up photo gives the model your real sleeves and arms instead of making them up.
Dress for the green. The notes are pale green grey, so a dark top, a blazer or a smart dress reads well behind them, while green clothes can blend into the cash. Avoid sunglasses, masks and hats, because the money crosses your face twice and a clear face is what lets your likeness return afterwards. Warm indoor light matches both rooms, while harsh flash or midday glare can look pasted on.
Accessories come from the photo as well. The model is told not to add jewellery or watches that are not in the picture, so if you want the joke about being so rich that you wear two watches, wear two watches in the photo. Likeness is never guaranteed, and if the first result does not look like you, change the photo rather than repeating the same one.
How do I choose between the two rooms?
The page offers two settings and the choice changes the mood more than you might expect. The evening bedroom is the one from the original clip: warm cream walls, a dark wooden bed and dresser, small plates on the wall and a ceiling fan turning at the top of the frame. It feels like a private, slightly cheeky flex, the kind of thing someone films at home for friends.
The chandelier living room is brighter and more obviously wealthy, with cream sofas, a large painting, sheer curtains and a sparkling light overhead. It reads as the polished, aspirational version of the joke.
How is this different from the CapCut template version?
Many of the earliest clips spread as a CapCut template, which is a ready-made edit where you place your own media into a fixed layout. Templates change or disappear often, so check inside CapCut for what is available today. videosi is independent and not affiliated with TikTok or CapCut.
A generated clip works differently. Instead of laying your picture over an existing edit, the model builds a new video in which you are the person hugging the cash, sweeping it past the lens and fanning it at the camera. Your face, outfit and arms are placed into a motion clip that supplies each move. That is closer to a short film of you than to a filter on a photo.
What does the video model actually do with my photo?
The clip is made with Seedance, a video model from ByteDance. ByteDance Seed describes Seedance 2.0 as accepting text, image, audio and video inputs, with images, audio and video usable as references to guide performance, lighting, shadows and camera motion.
In this trend that means two inputs do two separate jobs. Your photo decides who is in the picture and what they wear. A short reference clip decides the camera height, the hug, the scarf, the sweep, the flip and the fan. The text part is a five-part structure: an identity lock, the motion reference, the room and money, a timeline and a list of rules. We do not publish the full wording, but the structure is the reason every version feels like the same joke with a different person.
Do I need to label the video as AI when I post?
Yes, and it is the right thing to do even where it is not strictly required. TikTok's community guidelines on integrity and authenticity say creators must label AI-generated or significantly edited content that shows realistic-looking scenes or people. They also say unlabeled content may be removed, restricted or labeled by the platform, depending on the potential harm.
A money flex that looks filmed is exactly the kind of realistic scene that confuses people, so switch on the AI label or toggle in the app you post from. It costs nothing and it signals that the clip is a joke. A caption that says the cash is not real is a good second layer, particularly because the video is silent and viewers cannot be told by a voice.
Can I use this clip to look rich or to sell something?
No. The trend is a joke about pretending to be rich, and it only works while everyone knows it is a joke. The moment the same clip is used to suggest real wealth in a sale, a loan application, a dating profile or an investment pitch, it becomes deception.
Money flex videos are also a familiar tool of fraud. The Federal Trade Commission's page on investment scams warns that ads can promise a quick, easy, low-risk way to make a lot of money, that testimonials may not be genuine because the people giving them could be paid actors, and that anyone who promises a guaranteed return at low or no risk is a scammer. A video of a table buried in cash is easy to fake, so treat any clip like this as entertainment and never as proof of income.
The rule we follow is simple. Make it for fun, label it as AI, and never attach a pitch, a link to an offer or a promise of returns to it. If you see someone else using a clip like this to recruit people into an investment, do not join in.
What does it cost, and what happens if it fails?
A standard AI rich video (Standard model, 720p, 14 seconds) costs 12 credits. For comparison, a standard 12-second clip costs 10 credits, so the extra two seconds add a little. Better models and higher resolutions cost more.
There is no free generation and no free trial. Credit packs start at $6.99 for 10 credits and never expire, and subscriptions start at $29.90 a month. A failed, timed-out or blocked render returns its credits automatically.
A render that finishes but looks wrong, such as a fused fan or a face that drifted, is a normal result and the credits are spent. That is why the photo is worth a minute of care before you press Generate. The finished file is a silent MP4 with no watermark, so add the sound you want inside the app you post from.
When you are ready, the AI rich video trend page has the generator and example clips.