Make a Tiny Desk AI Video from a Photo
Make a Tiny Desk AI video from a photo by uploading one clear adult portrait to the original stage. The result is a 15-second, 720p vertical concert video with original stage audio. The solo pack is $9.99 for one video credit, paid once with no subscription.
Upload your photo and make a video →
Start with a photo that gives the face room
Choose one consenting adult with visible eyes, even lighting and an unobstructed face. A near-front view is a useful starting point. Avoid group shots, sunglasses, heavy filters and crops that cut off facial features. You need permission to use both the image and the person’s likeness.
Upload a JPG, PNG or WebP file up to 8 MB. If the only photo you have is blurry or the face is tiny, choose a clearer reference before buying a take. The photo checklist covers framing and lighting in more detail.
See how one portrait becomes a performance
The sample uses a fictional adult portrait and the original stage. Your uploaded photo guides the lead performer’s appearance. It does not supply a voice recording, replace each backing musician or guarantee identical clothing and expressions. Compare the actual input photo and output before choosing your own reference.
From upload to your first take
- Listen to the original stage. This song and singing voice stay in the result.
- Upload your portrait and continue with Google. No performance clip is needed for the original stage.
- Confirm permission and agree to the terms. Check your balance or buy a one-time pack on the pricing page.
- Submit one generation using one credit. Keep the page open while the photo uploads and the request is saved.
- Allow around 4–10 minutes, sometimes longer. Check My videos if you leave after the request is accepted.
A confirmed failure returns the reserved credit. A request still processing or being checked is not a confirmed failure; check its status before starting another take.
Check the full 15 seconds before sharing
- Watch the face at the start, during head turns and at the end. Does it remain recognizable?
- Inspect the hands, microphone and other performers for changes that distract from the result.
- Listen to the full soundtrack and compare mouth movements with the vocals. Keeping audio does not guarantee exact lip sync.
- Check the vertical framing on the device where you plan to share it. The output is 720 × 1280, not a landscape or 4K export.
Download with sound, or keep a silent copy
Download the regular MP4 to retain the original stage audio. “Download silent MP4” removes audio from that result without another generation. Adding a different track in an editor does not automatically align the lips to the new song. A different photo will not change the stage’s singing voice.
The stage is an original synthetic performance. It does not include the “Outstanding” song, NPR concert footage or a license to use either. This site is independent and not affiliated with NPR or any artist. Share the output as an AI-made video with media you have permission to publish.