How to Make a Chat GTT Video: The Complete Tutorial, From Script to Posting

How to Make a Chat GTT Video: Three Methods Compared

All three end in the same vertical clip for TikTok, Reels or Shorts. They differ in what you bring and who has to act.

The methods combine well: render a template clip, then join it with a reaction shot you filmed in an editor such as CapCut.

The Chat GTT Script, Beat by Beat

Whichever method you choose, the joke runs in the same order. Use this table as your script, your shot list or the outline of a prompt.

Transcripts online word the lines a little differently. The original video page lists the variants and explains where the sound came from.

The Whole Workflow in Seven Steps

These steps apply to every method. Where the methods differ, the step says so, and the sections below go into detail.

1. 1

   Cast the two roles

   Decide who is the cashier on screen and who is the customer holding the camera. In the photo templates, the face from your photo, or your pet, is always the cashier.

   Watch out: Whoever wants to say "I heard you the first time, bro" is the customer, which means filming and staying off screen.

2. 2

   Learn the script

   Go through the five beats in the table above: the decline, the identical repeat, the roast, the name-drop and the ending.

   Watch out: The repeat matters most. Practise saying it on one note, at one speed.

3. 3

   Make the picture

   Upload a photo to the Lookin' Ahh or Dog template, film the scene with a friend, or run a prompt in your own image-to-video tool.

   Watch out: For a photo, pick one subject facing the lens in even light. For filming, frame the cashier from the chest up.

4. 4

   Settle the sound

   Keep one audio source: the trend's original audio that a template clip already carries, the TikTok sound you recorded over, or your own voices.

   Watch out: Never layer the trend's sound over audio that already has the lines, or the words echo.

5. 5

   Trim and caption

   Cut dead air at the start, end right after the final comeback and add on-screen captions for the lines.

   Watch out: Many viewers watch muted, so captions carry the joke when the sound is off.

6. 6

   Label it

   Turn on the platform's AI-generated content label for a template or AI-tool clip.

   Watch out: A version filmed with real people and no AI does not need the label.

7. 7

   Post with the right words

   Put "Chat GTT" and "I heard you the first time, bro" in the caption, add a few tags such as #chatgtt and #carddeclined, and post vertically.

   Watch out: A short credit such as "inspired by @hangtimeindytv" points viewers to where the joke started.

Method 1: Make It From One Photo

This is the route with the least work, and the only one where a dog or cat can play the part without any editing. The template holds the streetwear store, the counter, the phone call, the dialogue and the trend's original audio. Your photo only decides who stands behind the counter.

Use the Chat GTT Lookin' Ahh template for yourself or an adult who agreed to appear, and the Chat GTT Dog template for a dog, cat, rabbit or other pet. Choose a sharp picture with one subject, the face toward the lens and nothing over the eyes, mouth or muzzle. A pet held in someone's arms is fine, because the pet template keeps only the animal.

Before the render starts, the button shows how many credits the clip uses. Credits come from one-time packs that never expire, and a failed render returns its credits. When the clip is ready, watch it once: if the likeness is off, a different photo usually helps more than rendering the same one again. Finished clips also stay in My Creations, so you can download them later. The I heard you the first time bro AI video guide has more photo tips for pets.

Method 2: Film It Yourself With a Friend

Two people, one phone and a spot that passes for a shop counter. These are the details that separate a clip people rewatch from one they scroll past.

• Set up a counter. A desk, kitchen island or shelf with a few products behind it is enough. Use a card reader, a calculator or a spare phone as the terminal, and give the cashier a second phone to hold to one ear.
• Light the cashier's face. Face them toward a window or a lamp. Even light on the face matters more than the set behind it.
• Get the sound, or skip it. On TikTok, open a Chat GTT video, tap the sound name and choose Use this sound, then mouth the lines in time while you record. If you would rather speak the lines live, skip the sound and rehearse the robot voice.
• Frame it vertically from the customer's eye level. Aim across the counter so the cashier fills the middle of the frame. Step closer instead of zooming, and keep the phone steady.
• Shoot the whole exchange in one take, several times. Three or four full takes beat cutting it into pieces. Keep the take where the cashier stays blank the longest and the roast has a clear gap before it.
• Edit lightly. TikTok's own editor is enough to trim and caption. If you want preset cuts and slots, the Chat GTT CapCut template page compares that route with the photo templates.

Method 3: Use an Image-to-Video Tool You Already Have

If you already pay for an image-to-video tool, you can build the scene yourself: give it one photo as the cashier and a prompt that describes the store, the phone call, the deadpan repeat and the customer who is heard but never seen. The Chat GTT AI prompt guide has copy-paste prompts for a person and a pet, explained line by line, so they are not repeated here.

This is the route for versions the templates do not make, such as your dog saying the comeback instead of the decline, or two friends swapping roles. Expect retries: general video tools often lose the robot voice or shuffle the line order, so check every render against the script table above.

Most of these tools generate their own sound. To use the trend's audio instead, mute the generated track and add the sound when you post, then make sure the lines still match the picture.

Common Mistakes and Quick Fixes

Almost every Chat GTT video that falls flat has one of these problems.

• The repeat sounds different. The second "the card declined" has to match the first. Reshoot with the cashier saying both on one note, or check your AI render against the script.
• The cashier reacts. A smile or an eye-roll breaks the robot joke. Reshoot, or for a template clip try a photo with a more neutral expression.
• The roast steps on the repeat. Leave a breath of silence after the second decline so the comeback lands on its own.
• The face is too small. Frame the cashier from the chest up, or upload a photo where the face fills more of the picture.
• The clip runs on. End on the customer's last comeback; anything after it reads as dead air.
• The sound is muted after posting. The platform did not clear the audio you uploaded. Post again without that track and add the trend's sound from the app's own sound library.