All effectsTalk to Me Doja
Talk to Me Doja: the cheek sway, made with AI from your photo
Talk to me Doja is a TikTok meme in which someone plays it cool through the build-up of a sped-up Doja Cat track, then puts both hands on their cheeks and sways like a shy anime girl when the beat drops. One photo becomes a 13-second vertical video of you doing it in a bright room at home, with the original sound; there are duo and pet versions too. From 7 credits.
A clear, front-facing face: the second half is all expression.
Video settings
This video: 7 credits. Ready in about 5 to 15 minutes; failed generations return your credits. View pricing
Updated
Your video
Your video shows here when it is ready, and stays in My creations.
Add your photos to start. Your video will show in this frame.
The video
What the video looks like
For the first seven seconds nothing much happens, on purpose. Someone stands in front of the camera acting unbothered: a scratch of the head, a shrug, a shift from foot to foot. Then the beat drops and they melt. Both palms go flat on the cheeks, the head tilts, the eyes close, and the whole upper body sways from side to side with a blissful smile until the clip ends. The less likely the person, the better it works.
The post this effect follows is the TikTok by @twin.cities.mattress, posted on 10 November 2026: a mattress store owner doing the sway in his showroom. Open it on TikTok to watch it. It is linked rather than played here because it carries the creator's caption and, from the seventh second, cartoon characters that are other people's art. No sample made on VideoEffect exists yet. Your video is not set in his showroom: the effect films the same build-up and sway in a bright room at home.
Second by second
Build-up and drop, second by second
Read from the original clip, one frame per second.
| Time | What is on screen |
|---|---|
| 0:00 to 0:02 | You stand facing the camera and scratch the back of your head. |
| 0:02 to 0:04 | A raised fist, then a shrug. |
| 0:04 to 0:07 | You shift your weight and open both hands, waiting. |
| 0:07 to 0:08 | The beat drops: both palms land on your cheeks. |
| 0:08 to 0:13 | Head tilted, eyes closed, you sway from side to side with bent knees and a huge smile. |
In the original post, pictures of the anime girl the pose comes from pop up in the corners during the sway. The video made here has nothing layered on top.
How it works
How to make the talk to me Doja video
- Choose a version: you, a duo, or your pet.
- Upload one photo per person, or one of your pet, and choose a resolution. The clip is always 13 seconds in vertical 9:16.
- Press Create and download. The price is on the button; the MP4 comes with the sound on it and no watermark.
Add the caption "talk to me Doja" yourself when you post. The video comes without text.
Posting it. The caption is the meme: type "talk to me Doja" over the video when you post, as the original takes do. The video comes with the sound on it and with nothing written on the picture. Having the sound on the clip does not give you rights to use it everywhere, so check a platform's rules before you put the video in an ad or a paid post, and switch on the AI-generated label when you post.
Versions
You, a duo, or your pet
You takes one photo: you play it cool for seven seconds, then your hands go to your cheeks and you sway.
Duo takes two photos, photo 1 on the left and photo 2 on the right. Both keep a straight face through the build-up and drop into the sway at the same moment.
Your pet sits upright on a bed, sniffs at the lens, then presses both front paws to its cheeks and rocks from side to side.
All three are set in a bright room at home, run 13 seconds and use the same sound.
Price
What 13 seconds cost
The video is 13 seconds, 9:16, with sound. The table below lists the credits for each resolution, from 7. The smallest pack is 10 credits for $9.99; pack credits never expire and a failed render returns its credits. This effect has no free preview.
| Length | 480p | 720p | 1080p |
|---|---|---|---|
| 13 seconds | 7credits | 12credits | 24credits |
The trend
Where "talk to me Doja" comes from
The sway is borrowed from an anime. In the first episode of "Yamada-kun and the Seven Witches", a character imitates the top student Urara Shiraishi with hands on the cheeks and a shy rock from side to side. On 7 November 2026 @lovingbalanci posted that scene with the caption "Speak to me Doja" over a sped-up cut of "Like That" by Doja Cat. Within days people were doing the sway themselves, and @twin.cities.mattress posted his on 10 November.
The meme goes by talk to me Doja, speak to me Doja, Doja talk to me and the Urara Shiraishi dance.
VideoEffect is not affiliated with Doja Cat or her label, with the makers or publishers of "Yamada-kun and the Seven Witches", with the creators linked here, or with TikTok or CapCut. No anime footage is used.
Photos
The photo behind the cheek sway
These points follow from the original clip; the effect has not been test-rendered.
- A clear, front-facing face. The second half is all about the expression.
- Knees-up or whole body. The clip frames the person from about the knees up.
- Hands down in the photo. The hands come up to the cheeks in the video and are drawn by the model.
- Clothes come from your photo. Dress as plainly or as seriously as you like; contrast makes the drop funnier.
- Use your own photo, or the photo of someone who agreed to be in the video.
Good to know
No GIF, no caption: what is left out
- It does not add the anime pictures or the caption that the original post has.
- It does not copy the showroom of the post shown above; the scene is a bright room at home.
- It cannot change the song or the timing of the drop.
- Two people at most.
- It is fixed at 13 seconds and 9:16.
Prompt
Copy the cheek sway prompt
Below are the three prompts, for one person, for two and for a pet. Each describes the room and the fixed camera, takes the face, hair and clothes from the photo, splits the clip into build-up and drop, and asks for a picture with nothing layered on top. No reference clip is used: the scene is written in words, and the sound is laid over the finished video.
You
A vertical 9:16 phone video, one continuous take of 13 seconds, real live-action footage, the camera a phone standing still at chest height, in a bright, tidy room at home with plain light walls, a bookshelf and soft daylight. A medium shot from the thighs up, the face large and clear in every frame. The person is the person in [Image 1] in every frame: face, hair, build and clothes exactly as in [Image 1], unchanged from the first frame to the last, in profile, with the head tilted and when the hands are on the face too. What happens: 0 to 7 s, the build-up: the person plays it cool: a lean toward the lens, a look to one side and the other, a step back into the room, arms loose, a small nod. 7 to 12.5 s, the drop: at 7 seconds both hands come up at the same moment and the palms land flat on the cheeks, fingers pointing up beside the face; with a shy, giddy smile and narrowed eyes, the head tilts from side to side and the upper body sways left and right, knees bending in time. 12.5 to 13 s: the hands come down. The video is filmed without speech. The trend sound is laid over the finished video afterwards, so every beat lands on the second given. Only this one person. Nothing is layered on the picture: no captions, text, cartoon or anime pictures, stickers, logos or watermarks.
Duo
A vertical 9:16 phone video, one continuous take of 13 seconds, real live-action footage, the camera a phone standing still at chest height, in a bright, tidy room at home with plain light walls, a bookshelf and soft daylight. Two people stand side by side in a medium shot from the thighs up, both faces large and clear in every frame. The person on the left is the person in [Image 1] and the person on the right is the person in [Image 2], in every frame: each keeps the face, hair, build and clothes of their own photo, unchanged from the first frame to the last. The two faces are never mixed or swapped. What happens, both in sync: 0 to 7 s, the build-up: both play it cool: a lean toward the lens, a glance at each other, a step back, arms loose, a small nod. 7 to 12.5 s, the drop: at 7 seconds both put their palms flat on their own cheeks at the same moment, fingers pointing up; with shy, giddy smiles and narrowed eyes, their heads tilt from side to side and their upper bodies sway left and right together, knees bending in time. 12.5 to 13 s: the hands come down. The video is filmed without speech. The trend sound is laid over the finished video afterwards, so every beat lands on the second given. Only these two people. Nothing is layered on the picture: no captions, text, cartoon or anime pictures, stickers, logos or watermarks.
Your pet
A vertical 9:16 phone video, one continuous take of 13 seconds, real live-action footage, the camera a phone standing still at chest height, in a bright, tidy room at home with plain light walls, a bookshelf and soft daylight. The pet sits upright on a soft bed in the middle of the frame, facing the camera, in a medium shot with its face large and clear in every frame. The pet is the animal in [Image 1] in every frame: the same species, breed, face, eyes, ears, fur colors, fur length and markings as in [Image 1], with the natural proportions of that animal, also when its head tilts and its paws are on its face. It wears nothing unless [Image 1] shows it. What happens: 0 to 7 s, the build-up: the pet plays it cool: a lean toward the lens with a curious sniff, a look to one side and the other, then it sits back, front paws relaxed. 7 to 12.5 s, the drop: at 7 seconds it lifts both front paws at the same moment and presses the pads against its cheeks; with a happy face and its mouth a little open, its head tilts from side to side and its upper body sways gently left and right. 12.5 to 13 s: the paws come down. The video is filmed without speech. The trend sound is laid over the finished video afterwards, so every beat lands on the second given. Only this one animal, with animal paws and never human hands. No people. Nothing is layered on the picture: no captions, text, cartoon or anime pictures, stickers, logos or watermarks.
What is the talk to me Doja trend?
It is a TikTok meme from November 2026: you act cool during the build-up of a sped-up Doja Cat song, then put your hands on your cheeks and sway like an anime girl when the beat drops.
How do I make the talk to me Doja video with AI?
Upload one photo on this page, pick a resolution and press Create. A 13-second vertical MP4 with the sound comes back.
Is it free?
No. It costs credits, from 7, with no free preview. Credits return if the render fails.
What song is used?
A sped-up cut of "Like That" by Doja Cat. The video carries the audio of the TikTok this effect follows.
What anime is the sway from?
"Yamada-kun and the Seven Witches". The pose imitates the character Urara Shiraishi.
Is it "talk to me Doja" or "speak to me Doja"?
Both are used. The 7 November post was captioned "speak to me Doja"; the take this page follows says "talk to me Doja".
Does the video include the anime GIF?
No. The video shows only you. You can add stickers or a caption when you post.
Can two people do it together?
Yes. The Duo version puts two people side by side; both play it cool and both drop into the sway at the same moment. It has not been test-rendered yet.
Is the example on this page made with VideoEffect?
No. It is the TikTok by @twin.cities.mattress, shown with credit. This effect has not been test-rendered yet.
Can my pet do the cheek sway?
Yes. The pet version sits your pet on a bed, and on the drop it presses its front paws to its cheeks and rocks from side to side. It has not been test-rendered yet.
How long is the video, and what format is it?
Thirteen seconds, vertical 9:16, as an MP4 with the sound on it and no watermark: about seven seconds of build-up, then the sway.
Is the person from the original post in my video?
No. The talk to me Doja take shown on this page is only the example. Your video is filmed in a bright room at home, with you, the two of you or your pet.
More effects
More meme effects
If the cheek sway worked on you, these effects are built on memes too.
Talk to Me Doja
Make yours
The price is on the Create button before anything is charged, and a render that fails gives its credits back.


