DEV Community

Cover image for How to Put Yourself in the Hotel Lobby AI Meme (From Two Photos)
Put Me In
Put Me In

Posted on Fully Autonomous

How to Put Yourself in the Hotel Lobby AI Meme (From Two Photos)

Hotel Lobby is the two-person routine everyone is remaking with AI right now: two people stand under one hanging microphone in front of an orange wall and run a tight set of hand gestures, glances and turns, cut into a handful of shots. You can tell it's the bit within a second, which is exactly why it works with your own face in it.

This guide covers how to make one starring you and a friend: which photos to use, the three steps, and what you get at the end.

What you need

  • Two photos, one for each person in the scene. The person in the left photo plays the left performer, the person in the right photo plays the right one.
  • Both people's OK. Only use photos of yourself or of someone who agreed to appear.
  • A few minutes while the clip renders.

Picking photos that work

The template decides the moves and the camera. Your photos decide who's in it, so they matter more than anything else you do.

  1. One person per photo. Each photo plays one person. A group shot makes it hard for the model to know who you mean, so crop it down to a single person.
  2. A clear, front-facing face. Faces turned to the side, or hidden by hair, sunglasses or a hand, often come out looking less like the person. Straight on or slightly angled works best.
  3. Even light. Soft daylight beats harsh shadows. Avoid strong backlight that leaves the face dark.
  4. A normal photo file. JPG, PNG or WebP, each under 20 MB. An everyday phone photo is fine.
  5. Only people who agreed to appear. Photos of celebrities, public figures or anyone under 18 are not allowed and are rejected when review catches them.

The three steps

We'll use the Hotel Lobby template on Put Me In, which is built for exactly this routine.

The Hotel Lobby upload screen on Put Me In, with one slot for the left person and one for the right
The upload screen: one photo for the left person, one for the right.

1. Add your two photos

Drop one photo into the left slot and one into the right. If you want to trade places, there's a swap button, so you don't need to re-upload.

2. Pick the format and quality, then pay once

Choose 9:16 (vertical, for phones and stories) or 16:9 (widescreen). Then pick the quality:

Quality Price
480p $6.99
720p $12.99
1080p $26.99

Your photos are reviewed first, and you pay once, after they pass. A rejected photo is never charged. There's no subscription and no free tier: it's one payment per video.

3. Get your clip

Rendering takes a few minutes. You can close the page; the finished clip waits under My videos in your account. Download it within 30 days, because finished videos are deleted after that.

What the finished video looks like

Two AI-generated fictional people in a suit and a silver jacket performing the Hotel Lobby routine under a hanging microphone
AI generated. Both people are fictional.

  • A 15-second MP4, in the 9:16 or 16:9 format you picked.
  • The template's music, with no voice added.
  • An "AI generated" label on every video. It is not real footage of the people in the photos, and it shouldn't be passed off as one.
  • The routine's shots and pace, with the two people staying recognisably themselves in each cut.

Two honest limits: lips move naturally but don't match any words, and hands can occasionally blur in close-ups. If a render fails on our side, you get a full refund.

Two AI-generated fictional older people, a man in a tweed jacket and a woman in a pink tracksuit, doing the Hotel Lobby arm move
AI generated. Both people are fictional.

What's happening behind the scenes

This isn't a face swap or a camera filter. A filter changes a video you filmed yourself; a face swap pastes a face into someone else's footage.

Here the template works as a motion reference: it carries the staging, camera moves and timing of the routine. The video model (Seedance 2.5) then generates a brand-new 15-second clip in which the people from your two photos perform it themselves. That's why photo quality matters so much: the model is building each person from a single image.

Common questions

Can I use a photo of a celebrity or a fictional character?
No. Photos of celebrities and public figures are rejected. Use yourself and friends who agreed.

Can I make one with my kids?
No. Photos of anyone under 18 are not allowed.

Can I use a group photo?
Crop it so each photo shows one person. One photo plays one performer.

Do I need a subscription?
No. You pay once per video, from $6.99 at 480p up to $26.99 at 1080p. Nothing renews.

Does the video have sound?
It comes with the template's music and no voice. If you'd rather use a different track, you can add one in the app you post to.

How long do I have to download it?
Finished videos stay under My videos for 30 days, then they're deleted.

Which format should I choose?
9:16 if you're posting to a phone-first feed or stories, 16:9 if it's going somewhere widescreen.

That's it: two good photos, three steps, and a few minutes later you and a friend are standing under the hanging mic. When you're ready, you can start on Put Me In.


This post was written with AI assistance. Images labeled AI generated show fictional people.

Top comments (0)