First Time Fancy Cafe
Free previewPricingThe trendSign in

How to Make Your Own Fancy Cafe Video

Six steps. The exact prompt is right below — copy it, no DM needed.

Copy the prompt

The Six Steps

  1. Pick two selfies. Front-facing, one person per photo, two different people.

  2. Make the first frame in an image model that takes photos (we used Nano Banana). Run prompt 1 with your selfie, then prompt 2 with that result plus your friend’s selfie. Two steps, not one.

  3. Turn that frame into part 1: image-to-video, 9:16, 11 seconds, prompt 3.

  4. Grab the last frame of part 1. Take it about 0.3 seconds before the end so you get a clean frame.

  5. Turn that last frame into part 2: image-to-video, 9:16, 10 seconds, prompt 4.

  6. Join the two clips (CapCut is fine), add the sound, caption it “POV: your first time in a fancy cafe” and post.

The Exact Prompt

Four prompts, in order. Two for the first frame, two for the video.

Prompt 1 — you on the table (image model, attach your selfie)

Image 1 is a person. Create a realistic smartphone photo, vertical 9:16, of a bright minimalist cafe corner with smooth concrete walls, a long low concrete bench and a round concrete pedestal table, seen from a three-quarter side angle at eye level. The person from image 1 sits ON TOP of the round table as if it were a stool: bottom on the tabletop, knees bent, sneakers on, feet hanging down, holding an iced coffee, calm, face turned toward the camera. Keep their face, hair and skin tone exactly. Nobody else in the frame. No text.

Prompt 2 — add your friend (image model, attach the result + your friend’s selfie)

Image 1 is a photo of a person sitting on a cafe table. Image 2 is another person. Add the person from image 2 into image 1, standing on the far left of the frame next to the low concrete bench, just arrived, holding a hot coffee cup, face turned toward the camera, keeping their face, hair, skin tone and build from image 2 and casual everyday clothes. Change nothing else in image 1: the person on the table, their pose, the table, the bench, the walls, the light and the camera angle stay exactly the same. No text.

Prompt 3 — part 1, 11 seconds (image-to-video)

Fixed camera, no camera movement, one continuous realistic shot. For the first 3 seconds the person sitting on top of the round concrete table calmly sips their iced coffee while the person standing on the left looks around. Then the person on the left walks over, sits down normally on the low concrete bench right next to the table, and calmly sets their coffee cup down on the tabletop beside the person sitting on it, without looking at them. The person on the table glances down at the cup, then at the person sitting on the bench, then down at themselves sitting on the table, and freezes with an embarrassed face. Keep both people’s faces, hair and clothes exactly as in the first frame, faces clearly visible. Quiet cafe ambience sound, no music, no talking, no text.

Prompt 4 — part 2, 10 seconds (image-to-video, start from the last frame)

Fixed camera, no camera movement, one continuous realistic shot, continuing the same scene. The person sitting on top of the round concrete table realizes the mistake, freezes for a second, then slowly and awkwardly slides down off the table and sits on the low concrete bench, trying to look casual. They pick up their iced coffee and sip it as if nothing happened, glancing sideways. The other person keeps calmly drinking their coffee and never reacts. Keep both people’s faces, hair and clothes exactly as in the first frame, faces clearly visible. Quiet cafe ambience sound, no music, no talking, no text.

Using Higgsfield

Higgsfield has image-to-video too — paste prompts 3 and 4 there for steps 3 and 5.

Common Mistakes

  • Describing the scene in text only. Video models put the person on the bench, not the table. Start from a first frame where you're already on the table.
  • Asking for all 21 seconds at once. Most video models stop around 15 seconds. Split it where your friend sits down.
  • Putting both people in with one image prompt. The AI often swaps who sits where. Add one person at a time.

Skip the six steps — two selfies, done. Make yours →

Keep reading: What Is the Fancy Cafe Trend?

First Time Fancy Cafe

Independent fan-made AI video tool. Not affiliated with, endorsed by, or authorized by any named studio, network, or artist. We do not claim to be official.

Product

Make a videoPricingThe fancy cafe trendMake it yourself

Legal

PrivacyTermsReport abuse

English only. Photos deleted within 24 hours. Videos kept 30 days. hello@firsttimefancycafe.com