Ear prompter proficiency is a listed skill on casting profiles, and the usual route to it runs through equipment. An earpiece, a recorder, a loop worn under your shirt, and enough practice to deliver against a recording without sounding like you are chasing it. The kit is sold to working professionals and priced accordingly, which puts the skill behind a purchase you have to make before you know whether you can do it.
You can practise the skill first. A browser and one earbud is the whole setup.
The flow
Three steps, and the middle one is the one people miss.
Rehearse the whole draft. With ear prompting switched on, read your script through once, start to finish. TellyPrompter captures that read as a guide track: your own voice, your own cadence, your own pauses. This is not an optional warm-up. The track is made during the rehearsal and only during the rehearsal, and there is no other way to produce one.
Keep the track. At the end of the rehearsal you are offered the recording. Accept it and it is filed against that draft.
Take. Put in one earbud, start the take, and the track plays a little ahead of you while you deliver to the lens. Nothing on screen, nothing in shot, nothing between you and the camera.
What adaptive means here
The reason ear prompting is a trained skill is that a recording does not care what you are doing. It plays at the speed it was recorded, and if you stumble or breathe or think, it carries on without you. Actors spend weeks learning to hold a steady distance behind a tape that will not stop, which is the technique described in how to use an ear prompter.
TellyPrompter is listening to your delivery, so the track paces itself against you. Stall on a word and it holds. Skip a line and it seeks forward to meet you. It never runs more than a set lead ahead, and you choose whether that lead is a sentence or around eight words. Which one feels right is personal.
That removes the distance-keeping problem. It does not remove the acting problem. You still have to sound like the words are occurring to you.
What you need
A browser, one earbud or earphone, and a quiet room to rehearse in. No recorder, no loop, no moulded earpiece, no account, and nothing to install. The prompter itself works offline once installed as an app.
Ear prompting works on iPhone, iPad, Mac and Windows.
What it does not do
Four limits worth knowing before you build a workflow on this.
The guide track cannot be captured on Android. The platform's speech recogniser starves when the page is also holding the microphone, so capture is blocked before the device is opened. A platform constraint, not a setting.
Voice mode needs the Web Speech API, which Firefox does not have. Use Chrome, Edge or Safari.
It reads your whole draft in one voice. If you are working a two-hander, the cue lines are read the same way your own lines are. There is no separate handling for a scene partner. For single-voice work — industrial copy, corporate scripts, narration, spokesperson pieces, your own content — that is exactly what you want. For scene work with a reader, it is not, and the actor apps built for that job will serve you better.
The track is bound to the exact draft you rehearsed. Change a word and the track goes stale, and you rehearse again to make a new one. This is deliberate, because a track that has quietly drifted out of sync with the script is worse than no track at all. But if your pages are being revised nightly, factor it in.
It takes a few goes
Hearing your own voice a beat ahead while speaking is a strange sensation, and the first thirty seconds feel wrong to almost everyone. The adaptive hold takes away the hardest mechanical part, but the sensation still has to become familiar, and that takes repetition.
Do not judge it on the first read. Run a sentence or two, then a paragraph, then a page. This is how ear prompter classes teach it in any case: listening only, then listening while speaking, then listening while performing.
Your own voice, on your own device
The track is your voice, which is the point rather than a limitation. You are mirroring your own cadence, not adapting to a synthetic one, and the phrasing you found in rehearsal is the phrasing you hear in the take.
Your audio never reaches us: TellyPrompter uses your browser's own speech recognition and runs no speech service of its own. Worth stating precisely if you are working under an NDA or an embargo, as industrial and corporate copy routinely is — some browsers recognise speech on the device, while others, Chrome among them, send the audio to their vendor's servers under their own policy. The script, the guide track and the recording stay with you either way: they are written to your browser's storage and your downloads, and never uploaded to us.
Where the hardware still wins
A concealed rig is a different thing. If a client books you for a shoot and expects an invisible earpiece with nothing on the table, buy the kit, and the equipment guide covers what one is made of. This is not that.
What this is: somewhere to build the skill before you spend anything, and a way to shoot your own self-tapes and content with the script in your ear instead of beside the lens. The credential is the thing worth having, and you can find out whether you can do it this afternoon.
For the background on why ear prompting has always meant hardware and training, start with ear prompting explained.