Bring your people.
One clear photo per person. Put each person on the side you want them to appear. Photos stay in your browser until you choose to review.
Add one clear photo per person, left and right.
Original duo rap with generated voices and matching mouth movement. Or choose instrumental dance.
New templates: ChatGPT Millionaire AI (ask how to get rich in 30 days, read the answer, stare) · Can-Can Dance AI (high-kick centre stage in a can-can line) · Tiny Desk AI (you on lead vocals at a cosy office-desk concert) · Gangnam Style AI video (you in the centre spot of the night-terrace horse dance) · Babydoll dance AI (you front and centre of the Babydoll two-step line) · AI baby dance video (you as a dancing baby, from one photo of you)
Free 5-second preview · Full videos from $6.99 · No subscription


Original AI characters.
Same references as the video below.
Generated voices and mouth movement together. Instrumental dance is also available. No uploaded-song lip-sync editing or artist voice cloning. Other photos and generations can vary.
duet.room is a Hotel Lobby AI video generator: upload a photo of you and one of a friend, and it makes a video of you both rapping side by side in the orange room, each on your chosen side, with generated voices and mouth movement. The first 5-second preview is free (480p, watermarked, no account or card) and usually ready in about 5 minutes. Full 10- or 15-second 720p videos cost 35 or 50 credits; full videos from $6.99, no subscription.
Pick your people.
We'll take care of the setting.
One clear photo per person. Good light, face visible, adults only.
Your opening act
Your partner in good times
Rap uses the same short original verses shown in our examples and newly generated voices. Mouth movement is generated with the audio; it is not an editor for uploaded songs or cloned voices.
Photos stay in this browser until you click Review video. Then they're uploaded to private storage with location data removed.
One orange room.
Your two people.


Original AI-generated characters — not real people — so we can show inputs and outputs side by side.
Real Wan 3.0 outputs from the two AI character inputs shown above. September 30 rap tests, played with sound. These demonstrate this pair; other photos and generations can vary. The raw 5-second test is unwatermarked; customer free previews are watermarked.
Which photos work, and what can go wrong: read the Hotel Lobby AI video guide ↗ · Migos AI trend & song · Rap & lip sync explained · Migos AI video generator · Hotel Lobby AI free · Hotel Lobby AI generator online · Hotel Lobby AI template · The Hotel Lobby trend
| What to check | duet.room |
|---|---|
| Two separate photos, no group shot | Yes — one photo per person, placed left or right as you choose |
| Both people rap, not just one | Yes — two generated voices take turns, with mouth movement generated to match |
| What the free tier really gives you | A 5-second video (not a still image), 480p, watermarked, downloadable MP4 — one per person |
| Sign-up and card | No account or credit card needed for the preview; Google sign-in is optional |
| Price of a full video | 35 credits (10 s) or 50 credits (15 s), 720p, no watermark; from $6.99 for one full video, no subscription |
| Time to render | Preview usually about 5 minutes; full videos usually 5–10 minutes |
| Original Hotel Lobby song | Not included — original beat and verses; no artist voice or likeness |
| If it fails | Credits return automatically; a failed free preview can be made again |
Honest limits: there is one scene (the orange room), likeness and mouth timing can vary between generations, and the full video is a new generation that can differ from the preview. Check that each face stays recognisable when heads turn, that nobody swaps sides, and that each mouth moves on its own lines.
| CapCut / Dreamina templates | General video models (Kling, Hailuo, Runway, Vidu) | duet.room | |
|---|---|---|---|
| Best for | A quick edit inside an app you already use | Full creative control over scene, camera and style | This one format: two photos in, a duo rap in the orange room out |
| What you do | Find a template and fill it with your clips or photos | Write a prompt, set up both people, add audio and lip sync yourself | Upload two photos, pick left/right and sound, generate |
| Audio | Depends on the template | Usually separate from the video | Generated with the video: original rap or instrumental |
| Cost | Free tier in the app | Free trials, then plan or credits | Free 5 s preview, then credits from $6.99 |
One clear photo per person. Put each person on the side you want them to appear. Photos stay in your browser until you choose to review.
Sign in with Google (no password), then we upload your two photos privately and show the exact price, format, delivery time and deletion dates.
Your first 5-second preview is free. Liked it? Turn the same photos into a full video with credits. If a video fails our checks, the credits come back.
Your first 5-second preview is free, so you can see your duo before spending anything.
Then buy a credit pack once — no monthly plan.
It's a two-person performance in a simple orange studio with one hanging microphone. People also search for this look as a Migos AI video or Hotel Lobby AI filter. We make an original performance with your own two people — not a copy of the original clip or its recording.
Choose one clear photo of you and one of your friend (face visible, good light). Add them to the left and right slots in the studio above, pick original rap or instrumental dance, and create the free 5-second preview — no account needed. If you like it, turn the same photos into a 10- or 15-second 720p video and post the 9:16 MP4 to TikTok, Reels or Shorts. Step-by-step guide
Yes: on duet.room the free preview needs no account and no credit card (you can sign in with Google to keep it with your other videos). It is a real 5-second video (480p, watermarked) you can download. What's free and what costs credits
Each photo drives one performer, and the side you choose is kept. Two clear, different-looking photos work best; covered faces, extreme angles and very similar looking people make it harder. Likeness can vary between generations, which is why the free preview comes first.
The free preview is usually ready in about 5 minutes. Full 10- or 15-second videos usually take 5–10 minutes. You can close the page and find the result under My orders.
Yes. Original duo rap creates a new beat, two generated voices and mouth movement together, using the short original verses in our examples. Choose instrumental dance for music without vocals. Voices, timing and likeness can vary between generations. Every video passes an automatic technical check before delivery, and credits return if we cannot deliver one that passes. This is not an editor for uploaded audio, and it does not clone Quavo, Takeoff or another artist.
Yes — your first 5-second preview is free (480p, watermarked, one per person). It shows your two people in the orange room before you spend anything.One full 15-second video is $6.99, or $19.90 for 3 and $29.90 for 5 — one-time payments, no subscription. The $29.90 option works out to $5.98 per 15-second video.
One clear, well-lit photo per person, face visible, no sunglasses or heavy shadows. JPG, PNG or WebP up to 10 MB, at least 256px on the short side. Only use photos of adults you have permission to use.
They stay in your browser until you click Review. Then they're uploaded to private storage with location data removed. Unpaid photos are deleted after 24 hours; paid photos 24 hours after your video is finished; the video after 7 days. You can delete everything sooner from your order page.
If creation fails or doesn't pass our checks within 24 hours, your credits come back automatically. Unused packs can be refunded to your card within 14 days. If you get a video that looks wrong, report it from your order page within 48 hours.
No. Videos use an original score. duet.room is independent of the recording artists and the original performance.