How to Make the Outstanding Tiny Desk AI Video

To make the Outstanding Tiny Desk AI video you need one photo of yourself, or up to three if friends join you on sax and keys, plus a Google account and credits. Upload the photos in the generator, pick 6, 10 or 15 seconds, and you get a vertical 720p MP4 with the original live sound. The page shows an estimate of about 5 minutes for 6 or 10 seconds and about 6–7 minutes for 15 seconds; you don’t have to wait on the page.

What you need

• A Google account. Sign-in on Tiny Desk AI is Google only; there’s no email and password sign-up.
• 1 to 3 photos. JPG, PNG or WebP, up to 10 MB each, one person per photo.
• Credits. 6 seconds costs 6 credits, 10 seconds costs 13, the full 15 seconds costs 20. Packs start at $9.90 for 20 credits. There are no free credits.

Step by step: your Tiny Desk AI video

1. Open the generator. Go to the generator on the homepage and sign in with Google.
2. Upload photo 1, you as the lead singer. This photo takes Charlie Wilson’s spot, the laughing, singing close-up. On a computer you can tap “Upload from your phone,” scan the QR code and pick the photo on your phone.
3. Add your bandmates if you want. Photo 2 is the sax player. Photo 3 is the keys player, and its slot opens once photo 2 is in.
4. Pick a length. 6, 10 or 15 seconds (see below for what each one shows).
5. Confirm and start. Tick the box to confirm everyone in the photos is 18 or older and agreed to be in the video, then press the button. You can leave the page; the video is saved to your account.
6. Download or share. When it’s ready, download the MP4 or use Share to TikTok, which gives you a caption to paste.

Who plays lead, sax and keys with 1, 2 or 3 photos

• One photo: you play all three parts: lead singer, sax and keys. The backing singers stay as in the original.
• Two photos: photo 1 sings lead and plays the keys; photo 2 plays the sax.
• Three photos: photo 1 sings lead, photo 2 plays the sax, photo 3 sits at the keys at the front, in front of the backing singers.

The backing singers and the other musicians stay as they are in the original clip.

6 vs. 10 vs. 15 seconds: shots, photos and credits

• 6 seconds, 6 credits: only the lead singer’s close-up, laughing and singing. Only photo 1 appears.
• 10 seconds, 13 credits: the lead singer’s close-up, then the sax solo in close-up, then the keys player with the backing singers.
• 15 seconds, 20 credits: the full version. The keys player with the backing singers and the sax player behind, then the lead singer’s close-up, the sax solo, and the keys and backing singers again.

Every length is taken from the original clip, and each one comes with the original live sound of that part of the performance.

Photo tips

• One person per photo. One clear subject is what works best for each part in the band.
• Face clearly visible. Nothing covering it.
• Good light. A well-lit face works best.
• Head and upper body in the frame. That’s the framing that works best for every part.
• Your own photos, not screenshots. The video model’s content check can refuse celebrity faces and screenshots with another app’s watermark. If that happens, the credits for the video come back automatically.
• Pets? A pet works too and stays an animal, but this hasn’t been tested on this site yet.

Other ways people make it

• A face-swap app. You find and download the original clip yourself, give the app one face per performer, check the tricky frames (head turns, hands, the sax at the mouth) and make sure the sound is still there before you export.
• A general AI video tool. You write a prompt for each shot, render them one by one, cut them together in the right order and add the audio yourself.
• Tiny Desk AI. You upload up to three photos and pick a length. Each camera shot of the original is re-created separately by the Wan 3.0 video model, and our server joins the shots and adds the original live sound in sync. AI results vary from one render to the next.

Posting it

• Format: the video is a vertical 9:16 MP4 in 720p with no watermark, so it fits TikTok, Instagram Reels and YouTube Shorts as it is.
• Share to TikTok: the button gives you a caption to paste. Before you post, turn on TikTok’s AI-generated content label.
• Hashtags people use: #tinydesk #outstanding #charliewilson #tinydeskai #aitrend
• Be clear it’s AI. Say in your caption that the video is AI-generated, and only tag friends who agreed to be in it.

Tiny Desk AI video FAQ

How long does a Tiny Desk AI video take?

Render times haven’t been measured on this site yet. The page shows an estimate of about 5 minutes for 6 or 10 seconds and about 6–7 minutes for 15 seconds. You can leave the page; the video is saved to your account.

Can I make it with just one photo?

Yes. With one photo you play every part: the lead singer, the sax and the keys.

Can I put my dog in it?

A pet works too and stays an animal, but this hasn’t been tested on this site yet. Use a clear photo of one pet with its face visible.

Can I recreate the college football coaches version?

No. Photos of celebrities and public figures, famous coaches included, aren’t allowed. Put your own team, league or coworkers in the band instead, with their OK.

What if the video fails or I don’t like it?

A shot that fails is retried once for free. If it fails again, the credits come back automatically. If you aren’t happy with a finished video, email [email protected] and we refund the payment to your original payment method.

Is it free?

No. There are no free credits. Packs are one-time payments: Single is $9.90 for 20 credits, Trio is $19.90 for 60 credits, and Crew is $29.90 for 100 credits. Credits never expire. See the prices on the homepage.