AI TikTok Video Generator

Start from a format, not a blank page

The templates are the fastest way in. ASMR Food renders a food subject in glass or crystal being spread across toast, with the clinking audio that format lives on. Fantasy Bed shows an impossible bed made entirely of one material, with someone climbing into it. Doorbell Cam is fixed security-camera footage of an animal at a front door. Animal Rescue is body-camera footage of a stray being pulled out of trouble.

There are others: a renovation time-lapse from a fixed ceiling camera, an enamel box that opens onto a miniature city, doll-sized product unboxings, a spray that restores a surface, tiny animals tumbling out of a jar, glossy gel squeezed into shapes, a drawing that climbs off the page, a sketch rising out of an old book, a working miniature shopfront, themed nail art in macro, and a giant animal watching over a tiny village.

You type one line. For ASMR Food that is something like strawberries made of red crystal. Everything else about the shot is already specified.

How a template actually works

Each template carries a format brief that describes the camera, the lighting, the pacing and the sound in detail. Your subject is written into that brief for every scene, and each scene is generated as an independent video clip against the result.

That is why the format holds across a video instead of drifting. The brief restates the camera position and the look in every single scene prompt, which is what stops five scenes turning into five unrelated shots.

You choose how many scenes, from one to ten, and how long each runs, from five to ten seconds. Length is scene count times seconds, and that is exactly what you are charged for.

Clips with their own sound

Animated clips are generated with native audio, so the ASMR crunch, the doorbell chime, the rain on the bodycam lens and the squelch of a jelly object all come from the generator rather than a sound library.

On the templates, that audio plays at full volume and there is deliberately no narration and no music bed. For ASMR and unboxing formats especially, the sound is the video, and a music track over it defeats the point.

Captions built for a vertical frame

Where a video does have a voiceover, captions are word-synced and sit high enough to clear the TikTok interface. Two styles, with a choice of highlight colour.

Captions require a voiceover, which requires a script, so the narration-free templates do not carry them. That is by design: those formats read as found footage, and a caption track breaks the illusion immediately.

The rhyme format

Rhyme videos generate an original song from lyrics you write, then cut animated clips over it. You can also upload your own audio track of up to three minutes and have clips generated against it.

Scene count is derived from the song's length rather than picked, at roughly seven seconds a scene. Pricing is the usual per-second rate plus a flat 250 credits for generating the song.

What a TikTok costs

Animated videos are 35 credits per second in HD, 25 at Standard and 65 at Full HD. A 30-second HD video is 1,050 credits. The shortest useful video, five seconds, is 175.

The $19 plan's 3,000 credits a month covers about three 30-second animated videos, or roughly 30 one-minute image videos. Most people running visual formats sit on the $49 plan.