Kinetic typography
A kinetic typography video maker that animates the words you actually said.
Drop in a clip and your speech becomes full-frame animated type, timed to your own voice word by word. No script to paste, no keyframes to set.
Kinetic typography is type that carries the whole frame and moves as a composition, rather than subtitles sitting in a band at the bottom of footage. LumaCaption ships 47 full-frame kinetic typography scenes that build themselves from your own speech and its word-level timings — no keyframing.
Every other kinetic typography tool starts the same way — you type a script, pick a template, and the animation runs on a stock rhythm that has nothing to do with a human voice. LumaCaption works the other way round. You bring a video with someone speaking in it, and the words thrown across the frame are the words that were actually said, landing on the exact beat they were said on.
That timing comes from word-level transcription of your own audio, so the type hits with your delivery instead of drifting past it. There are 47 kinetic typography scenes — Stomp, Sweep, Tumble, Halo, Cyber, Monolith, Boutique and Masala among them — and switching between them is one tap, not a rebuild. If you have ever hand-keyframed sixty seconds of talking head in After Effects, you know that is an hour gone before you have even looked at the colour. This is the same finish without the timeline.
Animated from your voice, not a template
Word-level timing is read from the audio itself, so every hit, hold and cut lines up with how you actually said the line — pauses and all.
47 kinetic typography scenes
Stomp for punch, Sweep for momentum, Tumble for chaos, Monolith and Boutique for restraint, Cyber and Masala when you want the frame loud. Part of a library of 290+ styles, all unlocked from the first video.
The After Effects hour, skipped
No comps, no expressions, no keyframing each word onto the waveform. Pick a scene and the whole clip is animated at once.
Built for typography reels
Full-frame type rendered straight into vertical video — 1080p on the free plan, 4K on paid — so a typography reel posts to Instagram, TikTok and Shorts as one file.
Kinetic typography is not the same as animated captions
An animated caption is an overlay: it sits over footage, occupies a band, and is designed to stay out of the way of whatever is behind it. A kinetic typography scene is the opposite — the type **is** the shot. It sets its own size hierarchy, uses the full frame, and treats the footage as a ground rather than as the subject.
That difference decides where each belongs. Kinetic scenes are for the first three seconds, for a punchline, for a title card or a quote pulled from an interview. They are not for a two-minute explainer, because full-frame type that changes every beat is exhausting to read for longer than a moment.
Where the motion comes from
A hand-built kinetic typography sequence is keyframe work: every word placed, sized and timed by hand, which is why a fifteen-second one is an afternoon. The whole thing is mechanical, though — word timings come from the audio, and the size hierarchy can be derived from which words carry the sentence.
LumaCaption scores each line, promotes the word that carries the point, and lays the rest around it at supporting sizes, then drives every entry off the real timestamp of the word being spoken. You get the composition without placing anything, and correcting a word in the transcript re-lays the scene rather than breaking it.
Making it readable, not just impressive
Full-frame type fails in a specific way: it becomes a graphic that people watch rather than a sentence they read. Two habits prevent it. Keep one clear hero word per beat, so the eye always knows where to land first — a scene with three equally-weighted large words has no entry point. And let the composition rest occasionally; a scene that changes on every single word never gives the viewer the half-beat needed to actually take one in.
The other constraint is the crop. A composition using the full 9:16 frame is cropped to 4:5 in an Instagram feed and 1:1 on a profile grid, both from the centre. Anything that has to be read belongs in the middle.
Kinetic scenes vs caption overlays
| Kinetic typography | Animated captions | |
|---|---|---|
| Uses | The whole frame | A band over footage |
| Footage is | A ground | The subject |
| Best length | Seconds — hooks, titles, punchlines | A whole video |
| In the catalogue | 47 scenes (Big Type lane) | The other 293 presets |
| Reading load | High — use sparingly | Low — designed to disappear |
How it works
Drop in a clip with speech in it
A talking head, a voice note over footage, an interview — anything with a voice. Your video file is never uploaded; there is no upload step to sit through.
Switch to the Big Type lane
Every word is already written and timed. Scrub through the 47 scenes, set your accent colour, and star any word you want the scene to build around.
Export the reel
Burned-in type travels with the video, so it looks identical everywhere you post it. Changed your mind on the scene? Re-export costs nothing.
Common questions
Can I type a script instead of speaking it?
No — and that is the point. LumaCaption animates spoken words, using real timing pulled from a real voice, which is what makes the result feel performed rather than templated. If your idea only exists as text, record it as a voiceover over the footage you plan to post and drop that in.
Is this a genuine After Effects alternative?
For spoken-word typography, yes. You give up per-frame control and custom rigs; you get back the hour you would have spent keyframing each word onto the waveform, and you get consistent timing across a whole series instead of hand-tuning every clip. Serious motion design still belongs in After Effects — a weekly typography reel does not.
How much of a scene can I change?
Colours, fonts, position and the transcript itself, plus which word each line builds around — one keyword per line is picked automatically and you can star a different one. Re-styling and re-exporting are free, so trying all 47 scenes on the same clip costs nothing.
Does it work for text animation reels in Hindi or Hinglish?
Yes. 93 languages are supported, English, Hindi and Hinglish are tuned end-to-end, and the typography scenes render whichever script your transcript is in — with a companion typeface matched to the style you picked rather than one fallback font for every script.
