The product is organised as a catalogue of trends rather than a prompt box. Each entry describes a specific viral format — a nighttime roadside exchange, a phone-meme reveal, a two-person dance duet, a red-lit freestyle performance — and states how many photos it needs and roughly how long the output will be, generally between 11 and 30 seconds. The original clip's choreography, camera movement and dialogue timing are preserved; what changes is who is performing it.
Generation is a short three-step flow: choose the template, add one clear, well-lit photo per role (people, pets and illustrated characters are all accepted), then pick length and quality. Standard output is 480p for a cheaper first attempt, HD 720p costs roughly double. The finished video arrives with the reference clip's original audio already synced to the performance, so it is ready to post. Typical render times run from about three to eight minutes for a short opening and six to twenty-five minutes for a longer clip, and the tab can be closed because rendering continues server-side and the result is saved to your account.
Pricing uses one-time credit packs rather than a subscription, with Google sign-in required to buy credits or generate. New accounts receive no free credits, so it is a paid-by-default tool, though purchased credits remain valid for twelve months and any credits held during a failed generation — including moderation failures — are returned to your balance. A few templates extend the concept, such as one that also lets you supply your own background and vehicle instead of keeping the original setting.
