Graphics for talking-head videos, timed to the words

1. 01

   Record as you normally would

   A phone and good audio is enough. Cueits needs to hear you clearly, because everything it draws is derived from what it can transcribe.

2. 02

   Upload and correct the transcript

   Fix anything misheard. A wrong word is a wrong graphic and a reveal on the wrong beat, so this step is worth the thirty seconds.

3. 03

   Review the beats before rendering

   Cueits divides the video into beats and proposes a graphic for each. You can move a boundary, change the wording or swap the element before anything is rendered.

4. 04

   Export

   One MP4, graphic burned in, ready to post. A render is only spent when you press export.

You do not have to. In portrait, Cueits puts the graphic above you rather than over you: the frame is split, the panel takes the top 1056 pixels of 1920 and your footage is cropped to the bottom 864. Nothing covers your face, and nothing is placed on top of what you are showing the camera.

This is a deliberate choice rather than a limitation. A graphic that floats over a speaker has to be small enough not to obscure them, which caps how much it can say; a graphic with its own half of the frame can carry a real chart at a readable size.

Because a graphic that arrives late is noise and a graphic that arrives early is a spoiler. If a figure appears three seconds before you say it, the viewer reads it, finishes the thought themselves, and stops listening to the sentence you are still delivering.

Synchronisation is the whole claim of the product. Cueits transcribes with a timestamp on every individual word, so a reveal is anchored to the phrase that motivates it rather than to a grid of even intervals.

No. You pick one of thirty-one looks, and the look is a complete identity rather than a colour scheme — its own typefaces, its own palette, and between one and six signature elements that only it draws.

What you can change afterwards is the content: the wording of any line, the figure in any statistic, which element a beat uses, where a beat starts. What you do not have to do is lay anything out.

The catalogue holds 129 elements. Charts and figures for anything quantitative, comparison and pros-and-cons layouts, timelines and step sequences, definition and verdict cards, pulled quotes, terminal output, star ratings, conversations, and a set of shapes for videos that are stories or tutorials rather than arguments.

The planner picks per beat. A beat that states one number gets a figure; a beat that weighs two options gets a comparison; a beat that walks through a process gets a sequence.

Will it work if I move around a lot on camera?

The crop is fixed for the whole video and centred where a seated speaker's face sits, so a clip shot from a tripod works best. Cueits does not track your face frame by frame.

Can I use my own brand colours and fonts?

Not yet. Brand kits are built and not available — every video currently uses one of the thirty-one supplied looks.

Does it work in languages other than English?

Cued Animations are planned in the language of the transcript. Captions cover about fifty-seven languages; Japanese, Korean and Chinese are not supported yet.