Hotel Lobby AI Doesn’t Look Like Me: Causes and Fixes
When a Hotel Lobby AI video does not look like you, the cause is usually the photo: a small, turned or filtered face, a head shot that leaves the model to invent the body, or another person in the frame. The next most common causes are pieces of the template's performers showing up, and the two sides being the wrong way round. Each has a fix below, and a 3-credit test render lets you check the fix before you pay for a long one.
Here is what RapVideo does with your photos, a table that runs from the symptom to the fix, and what happens when a photo is refused.
Updated
What happens to your photo first
On the default video, each photo is first redrawn as a standing studio photo of the same person, and the video is made from that. You do not see the redrawn photo, only your original. So the video is only as close to you as the photo you start from: the face the model can read, the clothes it can see and the body it has to guess all come from what you upload.
The privacy policy says the same, and that photos are deleted 7 days after the render and not used to train AI models.
Test with a short render
Try your fixed photo
Add the new photos and check the price on the button. A short test costs the least; a failed render is refunded.
Your two performers
One person each. Left raps, right hypes.
Stage
Video & sound
First a free 2-second preview, one per account. The full 12-second video is 6 credits at 480p. View pricing
From what you see to the fix
| What you see | Likely cause | Fix |
|---|---|---|
| The face looks like someone else, or is soft | A small, turned, backlit or filtered face in the photo | Use an unfiltered photo facing the camera, in front light, with the face large |
| Clothes or a body you do not own | A tight head shot, so the model invented the rest | Use a waist-up or full-length photo of what you want to wear |
| A blindfold, a cape, gloves or other costume pieces from the template | The model copied the template’s performers into your video | Make it again; if it comes back, use 16:9 and a different photo, and tell Support |
| Someone else from the photo appears | A second person in the frame | Crop to one person per photo, or use One photo with two clear faces side by side |
| The wrong person raps | The left photo raps and the right one hypes | Swap the two photos, or use Hand the mic to the other one after a render |
| Both people look the same | The same photo was added twice | Use two different photos; identical ones are refused before any credit is taken |
The photo rules behind the first rows, with examples, are in best photos for AI rap videos.
About the template costumes
The default Hotel Lobby video follows a reference clip of two stand-in performers in the orange booth. On 17 October 2026 we found that some 16:9 results had picked up pieces of their costumes, and we changed the 16:9 and 4:3 videos to follow a clip of two people in ordinary clothes. We did not measure the vertical 9:16 video separately, so if you see costume pieces there, switch to 16:9 and make it again.
The result of a render varies from one render to the next, even with the same photos. If a video keeps the problem after a second try, tell us through Support and include the video: that is how these cases get fixed.
When a photo is refused or a render stops
- Refused when you add it: the message names the problem (too small, not a photo, too wide). Fix that one thing and add it again.
- Stopped by a content check, before or after the render: the credits go back. Use another photo without revealing clothing. In our logs for the first two weeks of October, about 1 in 16 Hotel Lobby renders was stopped this way.
- Any failed render: the credits go back to your balance on their own. A standard 12-second video takes about 5 to 15 minutes on Wan 3.0.
Check cheaply, then make the long one
Render a short test first: 5 seconds at 480p costs 3 credits, against 6 for 12 seconds at 480p. If the face, clothes and sides look right, make the longer or sharper one with the same photos. RapVideo gives each new account one free 2-second watermarked preview of its default Hotel Lobby video or the Bulin 47 Freestyle.
FAQ
Why does my Hotel Lobby AI video not look like me?
Usually the photo: a small, turned, backlit or filtered face, or a head shot that leaves the model to invent the clothes. Use a front-facing, unfiltered photo from the waist up.
Why is the person on the right doing something I did not expect?
The left photo raps and the right photo hypes. Swap the photos, or use Hand the mic to the other one after the render.
Why are there costumes or a blindfold in my video?
The model can copy the template’s stand-in performers into the video. 16:9 now follows a clip of two people in ordinary clothes. Make it again, and tell Support if it comes back.
Why was my photo refused?
It was too small, not a photo, or stopped by a content check. Use another photo without revealing clothing. A refused or failed render is refunded or never charged.
Do I get my credits back if the video comes out wrong?
Failed and stopped renders are refunded on their own. A render that finishes is charged, even if you do not like the look, so test with 5 seconds at 480p first.
Can I use a screenshot of myself?
An original from the camera roll works better. Screenshots and photos passed through chat apps are often compressed and soft, which makes the face harder to read.
Read next
Hotel Lobby Trend With Your Dog or Cat: How to Do It
How to do the Hotel Lobby AI trend with a dog or cat: photos that work, who raps, one real test with a golden retriever, the cost and what to check before you post.
Best Photos for AI Rap Videos: What Works, What Fails
What makes a photo work or fail in the rap video generator, from the hard limits to light, angle, framing, sunglasses and group shots, with a fix for each.
Hotel Lobby Trend: How to Do It in Minutes From Two Photos
Three ways to do the Hotel Lobby trend (two photos and AI, a CapCut template, filming it yourself), with steps, free options and posting tips.
RapVideo is an independent tool, not affiliated with Quavo, Takeoff, Migos, Quality Control Music, Motown or COLORS.