The short answer
Migos AI Video at migos-aivideo.net is an independent two-photo Hotel Lobby AI video generator. Upload one photo per performer to create a vertical orange-studio duet with the preset reference audio. Each video is 15 seconds at 720p in 9:16 format and uses 50 site credits. Sign in with Google and buy a one-time credit pack to generate.
Check your two photos before generating
You need one image for each performer. Separate photos are fine; the two people do not need to be photographed together.
| Check | What to prepare |
|---|---|
| File format | Use JPG, PNG or WebP. A file must contain the image type it claims to be; renaming another file’s extension does not make it an accepted photo. |
| File size | Each image must be nonempty and no larger than 10 MB. The browser must be able to open it. |
| Subject clarity | Choose one identifiable subject per image, with visible facial features and even lighting. Avoid group shots, covered faces and tiny faces. |
| Framing | Leave space around the head and use similar camera angles where possible. A small image may be enlarged for submission; that cannot restore missing detail. |
| Permissions | Use images you own or are allowed to use, with the people pictured on board with the edit. Pet and character inputs may produce different results from portraits. |
An accepted upload is not a promise of a successful or accurate render. The generation provider also reviews inputs. If a file cannot be opened or submitted, follow the message in the studio and the photo troubleshooting guide.
Two performers, one preset reference performance
Your first photo identifies the left performer; your second identifies the right performer. They are two subject references, not the first and last frames of a transition between pictures. Use Swap positions if they are in the wrong slots.
The studio supplies the orange-room performance reference that guides the duet’s staging, movement and timing. You do not need to upload a base video or write a prompt. The output is fixed at 15 seconds, 720p and 9:16; there is no stage, duration or resolution picker.
A basic face swap replaces faces within existing footage. This workflow generates new duet visuals using your two subject photos and the preset performance, then combines them with the original reference audio. It does not simply paste your two photos onto an uploaded video, and the generated performance can vary.
The public gallery contains third-party reference clips for inspiration, not results generated by this studio. Watching another clip does not select it as a new performance template.
For an actual result created by our team, watch the cat duet and its performance reference. The example page explains the output format, audio check and observed limitations.
Sign in, generate and download your duet
- Sign in with Google. Open the studio and sign in to attach your credits and generation task to your account.
- Buy enough site credits. A standard video uses 50 credits. The Starter Pack costs $9.90 USD for 50 credits, covering 1 standard video. Packs are one-time purchases; applicable taxes may appear at checkout. See all credit packs. Credits arrive after payment confirmation.
- Select and review both photos. Assign the left and right performers and swap them if needed. Selection creates local thumbnails; both files are sent when you click Generate.
- Click Generate and follow the task. The studio prepares and reviews the photos, generates the visuals, then restores the reference audio. Pausing the browser’s wait does not cancel the submitted task. Resume that same task after a pause or refresh instead of submitting it again. A definite failed generation returns its reserved credits.
- Play and download the completed video. Results are available to the signed-in task owner. Your generated video is not automatically published in the gallery. Use the download button to save the MP4, then review it before sharing.
For a pending task, missing credits or a download problem, use the support guide. Do not start a second generation to recover a result that is still being processed.
Review the audio and result before sharing
The final video preserves the preset reference audio. It does not generate a new song or imitate a voice from either photo. Check both performers’ likeness, visible artifacts, movement and mouth timing: exact identity, choreography and lip synchronization are not guaranteed.
Review your rights to the audio and images before publishing. If you add music from a social platform’s library, check the options for your account and intended use. Uploaded photos and results are stored privately; inputs and task records have no automatic expiry in this version. Read the privacy policy for storage and access details.
Common questions about making a Migos AI video
What photos should I upload for a Migos AI video?
Use two separate JPG, PNG or WebP files, each no larger than 10 MB. Choose one clear subject per image, with a visible face and even lighting. The first photo is the left performer and the second is the right performer.
Are the two photos a first and last frame?
No. They identify two different performers who appear together in the duet. They are not the opening and closing frames of a transition between pictures.
Do I need to write a prompt or upload a base video?
No. The studio supplies the orange-room reference performance and its preset audio. You provide the two performer photos. Duration, resolution and framing are fixed at 15 seconds, 720p and 9:16. Watching a gallery example does not select a different performance template.
Can I generate a Migos AI video for free?
Selecting photos and browsing examples are free. Generating a video requires Google sign-in and 50 purchased site credits. The Starter Pack costs $9.90 USD for 50 credits; applicable taxes may appear at checkout. Definite failed generations return their reserved credits.
Does the result create a new song or voice?
The service generates new visuals and combines them with the original preset reference audio. It does not create a new song or voice from your photos. Exact lip synchronization, identity and movement are not guaranteed.
Are my uploaded photos and generated videos public?
No. Selecting photos creates local browser thumbnails. Clicking Generate uploads both files for private storage and video generation. Completed results are limited to their signed-in owner and are not automatically added to the public reference gallery. Inputs and task records have no automatic expiry in this version.
Sources and further reading
- Quavo & Takeoff’s original Hotel Lobby performance on COLORS: the visual inspiration for the orange-room duet trend.
- Current Migos AI Video pricing: one-time pack amounts and the standard per-video credit cost.
- Studio support: photo, task, payment and download troubleshooting.
- Privacy policy: when files leave your browser and who can access your results.
For the trend’s background, see What is Hotel Lobby AI? This independent tool is not affiliated with Migos, COLORS Studios or Higgsfield.