Pick a position on the video
Tap Lead, Sax, Keys, Back left, or Back right directly on the full reference image.
How to make a Tiny Desk AI video
The workflow is simple: identify the stage positions, upload clear photos, confirm permission, and review the complete fixed-template result.
The guided workflow
Each step removes one source of guesswork, from the face reference to the final export.
Tap Lead, Sax, Keys, Back left, or Back right directly on the full reference image.
Add a readable front-facing image for each person or pet you want to replace.
Confirm permission, create one complete video, and review the result in your private history.
Useful details
Tap Lead, Sax, Keys, Back left, or Back right on the reference image. You do not need to replace the entire group.
Use one clear adult or pet photo for each selected slot. Review the role label and thumbnail before continuing.
The result keeps the reference performance structure. Check the cast mapping, output quality, and AI disclosure before sharing.
Ways to use it
Use the smallest cast that answers your question, then add more positions when the full-group result is important.
Replace Lead or another single position before committing to a full cast.
Choose two or three adult friends or pets for a focused group-chat result.
Select all five positions and review every photo-to-role mapping before submitting.
What the generator controls
Your selected photos control the identity references. The fixed template controls the desk performance, camera timing, instruments, and audio direction.
Each uploaded photo is mapped to one selected position: Lead, Sax, Keys, Back left, or Back right.
The template keeps the reference desk, instruments, camera sequence, timing, and audio direction.
Choose one to five positions. Any empty position remains in the reference video.
Rights and retention
Only upload content you own or are authorized to use. Treat likeness consent, third-party rights, AI labels, and source retention as part of the production brief.
Generated tasks remain in your account history and use the shared media delivery rules.
See input guidance →How-it-works questions
Choose a performer position, upload one authorized photo, repeat for any other positions, confirm the consent checkbox, and submit the complete video request.
The configured workflow is designed to keep the reference scene, camera timing, instruments, and audio direction while changing the selected identities.
Select only Lead and upload one clear photo. Leave the other four positions empty so they remain in the reference.
Completed jobs follow the site’s private history and media delivery flow. Check the result state or history entry after submission.
Check the original task status and failure message before submitting the same request again. Qualifying failures follow the site’s configured credit-refund path.
Review the source rights, likeness permission, AI disclosure, third-party materials, and destination-platform rules before publishing.
Ready to follow the workflow?
Start with one position or prepare the full cast. The creator shows the mapping before you submit.
Keep your video moving
Choose a one-time pack and continue with the video template you already selected.
One complete cast experiment.
Compare three cast combinations.
A full set of creative cast tests.