UNPLUGGED AI VIDEO / The guide

How to make an AI birthday video
from one good photo.

Start with a clear portrait and a personal note. Here is what you can try today, and how the planned acoustic session will work.

01

Choose a photo that looks like them.

Use one well-lit image with the person facing the camera and their whole face visible. Avoid sunglasses, face-covering hands, strong beauty filters, deep shadows and group shots. An ordinary clear portrait is more useful than a heavily edited headshot.

For a pet, keep the eyes, muzzle and ears in focus. Avoid cropped ears, a hidden face or motion blur. One photo is enough to start the draft; a three-view character sheet is not required.

The current preview accepts JPG, PNG and WebP up to 10 MB, with at least 320 pixels on each side. Use your own photo, a consenting adult’s photo or a pet photo you have permission to use. Human subjects must be 18 or older.

02

Keep the dedication short and personal.

Add a first name of up to 32 characters and a greeting of up to 160 characters. Try a shared memory, a small joke or a wish for the year ahead. Names appear as text; singing a custom name is not included in the first session.

For example: “Happy birthday, Alex. Here’s to another year of road trips and very loud sing-alongs.” You can change the message freely in the local preview.

The birthday gift page explains the intended delivery and planned price.

03

Try the draft before there is a charge.

Open the photo preview tool, choose your image, add the dedication and confirm photo permission. Review the greeting and download it as a PNG image. The downloaded card contains your words, not your portrait.

This is a local draft. Your photo is not uploaded, put into a scene or sent for generation. The completed videos on the examples page are separate demonstrations. Closing or refreshing this draft clears its unsaved photo and text.

04

Know what the upcoming video service includes.

The planned order prepares a new scene with one lead, then generates an approximately 20-second video and its own singing. The acoustic trio stays in a living room; the two backing musicians are fictional. The output is a portrait 720p MP4.

The initial wait estimate is 7–10 minutes, with extra time possible for queues and checks. Once server-side orders are connected, leaving the page should not cancel a submitted task: you should return to the same task and download its result. This feature is not active in the current draft.

The launch rule is one credit per complete video, an upfront credit cost and an automatic credit return after a final generation failure. Reopening a task should not create another charge. Review the planned prices and credit-return policy.

Listen before choosing.

Play a real sample with sound to hear the singing style and acoustic arrangement. The voice is synthetic and does not recreate the person’s real voice. The four planned singing styles are adult male, adult female, boy and girl; each can be chosen for a person or a pet. The samples illustrate two results, not every voice combination. Uploaded recordings are a later feature.

Keep your original file.

The intended paid-service policy is to delete source photos and working copies within 48 hours after delivery, with finished-video storage described separately before checkout. For now, the local preview stores nothing after you leave. Read the current privacy notice.

YOU KNOW WHO THIS IS FOR.

A small stage.
A very personal moment.

Try the photo preview

Free local draft. Video orders and payments are not open yet.