"I thought we were having a nice date" is the kind of line that stops a scroll: said sweetly, then said again through tears. This guide covers where the meme comes from, why it works and how to make a 13-second version of the two of you from 2 photos with the AI I Thought We Were Having a Nice Date template on Viberot.

Two people at a dim restaurant booth, one standing with a sweet smile, the other looking up
The nice date moment

Where the meme comes from#

The meme comes from a restaurant date scene in the 2026 horror film Obsession. One person sits in a dim, wood-panelled booth giggling "No, no, no", stands up with a sweet smile and says "I thought we were having a nice date". The date across the table reaches out, "We are, we are, sit down", and the smile collapses into tears: "Why does it matter? I thought we were having a nice date!" The official clip took off in May, and the scene keeps coming back as a template for people remaking it with their own faces, which is why searches for the nice date meme keep popping up.

Why it works#

  • A big swing in a few seconds. Giggling, a sugary smile and a full breakdown, all in one booth.
  • Everyone knows the date that goes wrong. The line is funny because it is both polite and unhinged.
  • Great for reactions. The face change is the whole joke, so the clip needs no extra editing.
  • Works for any pair. A partner, a friend, a sibling or a coworker all fit the format.

The scene, beat by beat#

Here is what the 13-second version you get looks like:

  1. The giggle. Close-up of one person in a dim booth, trying not to laugh: "No, no, no..."
  2. The reverse shot. The date across the table, confused, half-eaten plates between you.
  3. The stand-up. A wide shot as they rise, smiling sweetly: "I thought we were having a nice date."
  4. The plea. The date leans in and reaches out: "We are, we are, sit down."
  5. The breakdown. The smile cracks and the tears come: "Why does it matter? I thought we were having a nice date!"

Each person keeps their own face, hairstyle and clothes from the photo. The clip comes with AI voices speaking the lines, lip-synced to both of you, with voices that suit each person.

Step by step: how to make it with AI#

  1. Open the template. Go to the AI I Thought We Were Having a Nice Date page and sign in.
  2. Upload two photos. First the person who breaks down, then the date sitting in the booth, both with the face visible.
  3. Pick the aspect ratio and quality. 9:16 for TikTok, Reels and Shorts, or 16:9 for YouTube. Standard 480p or 720p is enough.
  4. Generate. No prompt to write: the booth, the lighting and the acting are set. Rendering usually takes a few minutes.
  5. Download and post. You get a 13-second MP4 with the spoken lines already in it. Post it as it is, or add captions in TikTok.

Picking photos that work#

Works well Causes trouble
One subject per photo, face clearly visible Group photos where the faces are tiny
Front-facing, relaxed or neutral expression Extreme angles or a face turned away
Hair and shoulders visible for the one who breaks down Hats, masks or sunglasses that hide the face
Even light, sharp focus Heavy filters, very dark or blurry photos

Note

Photos that appear to show public figures, including actors, singers and influencers, are rejected before any credits are charged. Use photos of yourself or of friends who agreed to be in your video.

Ideas for your post#

  • The group chat reply. "POV: I said the plan was fine and then I read the group chat."
  • The couple version. Post it with your partner as the date, or tag the friend who always ruins dinner.
  • The petty grievance. Caption it with the smallest thing that has ever made you cry, like a cold fry or a late reply.
  • Pair it with another trend. Follow it with I Love the Smiths AI for a second awkward-sweet post.

How it differs from a CapCut or editing template#

A CapCut template asks you to film or find clips and drop them into fixed slots, so you still need a booth, a second person and a good crying take. Here you only upload two photos: the restaurant, the lighting, the camera angles and the acting are generated for you, and both faces and outfits carry over. The result is a short clip with the lines already voiced, ready to post as it is. The video has no film scenes or film audio, and Viberot is an independent tool, not affiliated with the film, its studio or distributor, or any of its actors.

Price and FAQ#

A Standard 720p video costs 600 credits and 480p costs 272; see pricing for credit packs. Credits are a one-time purchase, with no subscription.

Does the video include the film audio or the spoken lines? It includes the spoken lines in AI voices that fit each person. There are no film scenes and no film audio in it.

Can I use a pet? This template is built for people.

Can I use photos of the actors? No. Photos of public figures are rejected before any credits are charged.