The First Time in a Fancy Cafe AI prompt
The First Time in a Fancy Cafe AI prompt has to do five things at once: keep the table tiny, hold the coffee cup steady, keep you identical to your photo, make the newcomer a different face, and time the joke. On First Time in a Fancy Cafe AI it is already written and tested. Make the video from one photo.
What does the prompt have to solve?
- The tiny table: a round steel table small enough to look like a stool, fixed to a long concrete bench.
- The cup: a takeaway coffee that stays the same size from the first sip to the moment it is set on the table.
- Your face: the person in the video must stay identical to your photo, with your hair and clothes.
- The newcomer: the second person must be a new face, not you again and not the original creator.
- The timing: she sits, puts her cup down, he freezes, then climbs down and sits on the bench like nothing happened.
Why do typed prompts fail?
- The table grows. Models default to a normal cafe table, so the stool joke disappears.
- The cup changes. A small object in a moving hand is redrawn from frame to frame.
- Your face drifts. Without the right inputs the face and outfit change as the person climbs down.
- Two copies of you. Both people in the scene can get your face unless the newcomer is set apart.
- The beat is lost. A prompt describes what happens, not how long the pause lasts before he climbs down.
What is built in here?
Every template has a fixed, tested prompt and follows one shot list of 12–15 seconds, so every render goes through the same beats. You supply one photo and pick a template. The model is Seedance by ByteDance, and a standard 720p video costs 10 credits. Even so, the cup and the table can wobble on some tries, and a second render usually fixes it. See how it works.