A teleprompter that knows whether you'll run long

Most online teleprompters solve one problem: they keep your eyes near the lens while the words go by. Useful — but it is the second problem that ruins takes. You find out the script ran ninety seconds only after you said all of it.

That is the gap the timer coach closes. Pick a target of 30, 60 or 90 seconds and the prompter tells you, before you press record, roughly how long the full script runs at the speed you have set.

While you read, the elapsed clock runs against the target and the coach counts the words still to go, so drifting long is something you see in the moment instead of in the edit.

The arithmetic is deliberately simple. The scroll speed and the height of the text imply a runtime — distance divided by speed — and the word count divided by that runtime is the reading pace. Nothing you cannot check yourself.

It is honest about what it cannot know, too. The scroll cannot hear your pauses, so a deliberate beat adds seconds the estimate does not see. Land a touch under the target and you are covered.

The one mistake that makes prompter reads sound robotic is chasing the scroll. The text should follow your voice, not drag it.

Start with the readout rather than the stage. Move the speed slider until the estimated runtime sits where you want it, then do one rehearsal pass out loud and notice whether the cue line arrives with your voice or ahead of it.

If you know your real speaking pace, the guessing ends. Measure it once at the companion pace test at /tools/speaking-pace-test and this page quotes your spoken runtime beside the scroll estimate.

Text size is a distance decision, not a taste decision. Reading from a phone at arm's length, the smaller sizes work; reading from across a room, push it up until whole phrases land in a glance.

Bigger text also takes longer to scroll past at the same speed, so the runtime estimate moves when you resize. Check it again after any change — the readout under the slider recalculates live.

Put the stage as close to the lens as you can. On a phone that means this page scrolled so the stage sits at the top of the screen, right under the camera — the few degrees of eye-drop disappear at normal viewing distance.

Step back if you can. The farther you stand from the screen, the smaller the angle between the lens and the words, and the less your eyes appear to track the lines.

Write like you speak before you prompt a word of it. A script full of written-English clauses reads as recitation no matter how good the scroll is; short sentences you would actually say survive the trip to camera.

Rehearse once, record once. The first pass finds the tongue-twisters and the sentence that needs a breath; the second pass is usually the keeper.

The mirror toggle is for beam-splitter rigs, where a hood of angled glass reflects the screen in front of the lens. The reflection flips the text, so the prompter pre-flips it back. Reading straight off a phone or a monitor, leave it off.

The targets are not arbitrary — they are the lengths short vertical video actually gets watched at. Thirty seconds is one claim and one piece of proof. Sixty is the standard ad and explainer length. Ninety is about the ceiling before attention sags.

At a typical speaking pace of 130 words a minute, those targets translate to about 65, 130 and 195 words. That 130 is an assumption, which is exactly why the coach works from your actual scroll speed instead of an average.

Speaking paces genuinely differ: a careful reader at 110 words a minute and a quick one at 150 fit very different scripts into the same sixty seconds.

If the coach says you are long, cut before you speed up. A script forced to fit at a sprint sounds like the terms and conditions; the same script minus two sentences sounds like a person.

Cut in a fixed order: the second example first, because one example carries the point as well as two, then the warm-up opening, because the real first line is usually sitting in sentence three. The claim, the proof and the instruction stay.

A prompted read that lands on target is the hard half of a short video done. The other half — the cut, the captions, the b-roll — does not need a timeline.

Cutroom turns one spoken take of up to three minutes into a finished 9:16 video. The transcript drives the edit: captions land word by word, and clips cover the phrases you highlight.

The target discipline keeps paying after filming. If the take still runs long, delete a line in the transcript and the cut rebuilds around it — no razor tool, no ripple edits.

Change the pace of the edit and the machine re-cuts the whole thing, which is how one prompted script becomes a 60-second version and a tighter 45 in the same sitting.

The pricing is the same arithmetic as the coach. One batch is 100 credits and the export is 20 credits per output minute, so a 60-second finished video is about 120 credits.

Three finished videos are free inside your first seven days, no card, on the 600 credits granted at signup, and they export with a Cutroom mark across the middle. Lite is $19.99 a month for 1,250 credits — about ten of those finished one-minute videos, and more when they run shorter.