Loading…
Loading…
TIKTOK
Vertical 9:16, native and fast — hook in the first beat.
This AI TikTok video generator writes for how TikTok actually reads. The feed is native-first: polished, ad-shaped delivery gets scrolled past, and the most common way a generated TikTok fails is by sounding like a brand presenting rather than a person talking.
So the script opens mid-thought. Context can arrive a beat late; curiosity cannot. The rhythm is spoken rather than written — contractions, asides, direct address — and it is capped at under a minute, where the format works best.
Output is vertical 9:16 with burned-in captions, generated from a script rather than clipped out of something longer.
The single biggest tell of a generated TikTok is a run-up. A greeting, a channel name, a slow establishing line — all of it reads as an ad, and the feed treats ads accordingly.
Opening mid-thought is the fix, and it is built into how the script is written. Start at the interesting part and let the viewer catch up; they will, and they will stay while doing it.
Comment bait is a feature of the format rather than a flaw in it. A deliberate omission — one thing left arguable, one detail people will correct — reliably drives comments, and comments drive reach more directly on TikTok than on either of the other two platforms.
This is different from being wrong. The gap should be a genuine open question, not an error you are hoping nobody notices.
Captions are burned in, and the pacing is tighter than on the other vertical platforms. TikTok viewers scroll faster and forgive less, so the cut rate is higher and dead frames are shorter.
Under a minute is the working ceiling. Longer TikToks exist, but they need a much stronger reason to hold, and a generated one rarely has it.
TikTok has no tags field, so hashtags carry discovery — around a dozen, mixed between one or two broad and several specific to what the video is actually about.
The caption itself should be short. A single punchy line usually outperforms a paragraph, and a question in it reliably drives the comments that drive reach. The Caption and Hashtags tool generates both together.
TikTok distributes video by video rather than by channel far more than the other two platforms do. A creator with two thousand followers can have a video reach a million and the next one reach four hundred, and neither number says much about the account.
The practical consequence is that consistency matters more than optimisation. Anything that reduces the cost of making the next video — a script you did not have to write from nothing, a render you did not have to edit — is worth more than any amount of tuning applied to the last one.
It also means a flat video is cheap to be wrong about. Post it, look at the watch-through in the first two seconds, and change the opening on the next one. That loop is the whole strategy on this platform.
● QUESTIONS
Under a minute, which is what generation is capped at. Longer TikToks work, but they need a much stronger hold than a generated video usually has.
TikTok requires realistic AI-generated content to be labelled, and provides a label for it. A captioned video over stock footage with a synthetic voiceover is not the case that rule targets, but labelling when in doubt costs nothing.
Because a run-up is what gets a TikTok scrolled. Context can arrive a beat late; curiosity cannot. Starting at the interesting part is the single highest-leverage change in the format.
Yes. No camera and no on-screen presence is needed — the voiceover is synthesised and the footage is sourced or generated.
Around a dozen, mixed between one or two broad terms and several matched closely to the content. TikTok has no tags field, so hashtags are the discovery signal.
● OTHER FORMATS
Full-length 16:9, no time limit. Paste a whole script.
Vertical 9:16, under a minute, built for the Shorts shelf.
Vertical 9:16 with the longest runway of the three.
Ideas, scripts, thumbnails, SEO and the rest of the pipeline