How to Make a Tiny Desk AI Video from a Photo
Learn how to make a Tiny Desk AI video in your browser: upload photos, assign the lead and sax, and download AI music and silent versions. See real examples.
By Tiny Desk AI / Smart Lazy LLC · Updated
Made here, from these photos
Play with sound to hear the original AI music.
Cream jacket
Emerald blouse
Silver hair
Cat on lead · experiment
These 15-second videos were made with our generator using original AI portraits. The cat is an experimental example; pet results are not guaranteed. Results vary with each photo.
1. Choose your lead photo
Open the studio and add one clear adult portrait to Lead singer. JPG, PNG and WebP are accepted, up to 10 MB per image. Keep the face sharp and visible, and include the shoulders and the clothing you want the model to use.
The examples above use three different fictional adult portraits: a man in a cream jacket, a blonde woman in an emerald blouse and a silver-haired woman in a plum jacket. A fourth example uses an original orange cat portrait as an experimental lead. All four use the same 720p generation flow. Results vary; these examples do not establish reliable results for every face or pet.
2. Add a friend on sax (optional)
Only the lead is required. To make a duo version, add a second photo to Sax player; that person plays the sax solo near the end. Without it, the reference musician stays. Keys and the back row always stay as performed.
Each slot shows the performer it replaces, and the preview beside it jumps to that person’s shots. No prompt or mode is needed: the video uses each person’s face, hair and clothing, and the reference guides the performance and camera. Use 9:16 for a vertical clip. The current generation setting is fixed at 15 seconds and 720p.
3. Review the cost and generate
Sign in with a verified account and review the setup before confirming. One credit makes one video, whether you use one photo or two. The one-time packs are $5.90 for one video, $14.90 for three, and $39.90 for ten. There is no subscription, credits do not expire, and a confirmed failed generation returns its credit.
If you leave to sign in or buy credits, this browser keeps a temporary copy of the creation for up to 24 hours. Returning does not automatically start generation. Review the settings and confirm when ready; do not close the browser expecting a saved setup to contain a permanent copy of your photos.
4. Inspect and download the result
Open My videos to check progress and play the completed clip. Look at the lead close-ups, the wide shot and the sax solo if you assigned it. AI can change facial details, hands, clothing and motion. Our example demonstrates one outcome; it is not a guarantee that every portrait will match equally well.
Download the version with its AI-generated soundtrack or choose the silent copy. The included music is generated for the clip rather than a copy of the original Outstanding recording. If something went wrong, use the problem-report link on the video so support receives its task reference.
5. Add a platform sound when publishing
To use a trending Outstanding sound, import the silent download into TikTok, Reels or Shorts and choose a sound available within that app. Availability depends on your account and location; the generator cannot guarantee that a particular recording will be offered. Adjust where the audio starts and preview the whole clip: adding a sound does not automatically synchronize the cuts, gestures or lips.
Identify the clip as AI-generated, use photos with permission and avoid implying that the depicted people really performed or endorsed the video. You can share the generator link with your post so viewers know how your version was made.