How to use the voice teleprompter
- Paste your script. Press Edit script and paste or type it, or import a .txt, .srt or .vtt file. A ScribeCue transcript can be sent here with one click.
- Pick how it scrolls. Private voice runs a speech model on your device (English). Browser voice uses your browser's speech service and supports 20 languages. Auto-scroll needs no microphone.
- Set it up for your camera. Make the text bigger, move the reading line next to your lens, turn on Mirror for a glass teleprompter, and go full screen.
- Press Start and read. Allow the microphone. The next word to read sits on the red reading line, the text waits when you pause and catches up if you skip ahead.
How the text follows your voice
Every fraction of a second, ScribeCue turns the last few seconds of speech into words and lines them up against your script. It looks a little behind where you are and well ahead, allows for words the recogniser hears wrong, and prefers a short step forward. That is why it doesn't jump when you ad-lib, repeat a phrase or say a common word like "the" that appears all over the page.
- Pause and it waits. Nothing scrolls while you are silent, so you can breathe, smile or look at a guest.
- Skip a sentence and it catches up. It finds you again within a few words, up to about 80 words ahead.
- Go back and it follows — or click any word to jump straight there.
- Scroll by hand with the wheel or a finger; following pauses for two seconds and then picks up where you are.
Three ways to scroll — and where your audio goes
| Mode | Speech recognised by | Languages | Audio leaves your device? | Best for |
|---|---|---|---|---|
| Private voice (default) | A small speech model (Moonshine, ≈28 MB) running in this browser tab | English | No — not even to us | Scripts with sensitive content, offline-ish use after the first load |
| Browser voice | Your browser's built-in speech service | 20 languages | To Google (Chrome, Edge) or Apple (Safari), not to ScribeCue | Non-English scripts, slower laptops |
| Auto-scroll | Nothing — no microphone | Any | No audio at all | Firefox, very noisy rooms, music beds |
The model for Private voice downloads once from a public CDN and is cached by your browser. ScribeCue never receives your recording in any mode; your script is saved only in this browser's local storage so it is still there next time.
Set it up for a camera, a phone or a glass rig
- Text size from 20 to 120 pixels; ↑/↓ on the keyboard while reading.
- Reading line — move it up near a webcam so your eyes stay close to the lens.
- Mirror flips the text left-to-right for beam-splitter glass; Flip upside down for rigs that need it.
- Text width narrows the column so your eyes move less on a wide monitor.
- Full screen (F) hides everything else; the controls fade while you read.
- Countdown gives you three seconds to look up before the first word.
Keyboard: Space starts and pauses, ←/→ moves six words, M toggles mirror, F toggles full screen. A Bluetooth presenter that sends arrow keys or Space works as a remote.
How it compares with other free voice teleprompters
| ScribeCue | Speakflow (free plan) | Sozai | Smart Teleprompter | Teleprompter.com | |
|---|---|---|---|---|---|
| Voice scrolling | Unlimited | 10 min a month | Unlimited | Unlimited | Yes (VoiceGlide) |
| Sign-up | None | Account | None | None | Account / app |
| Speech recognised on your device | Yes (Private voice) | No | No — browser service | No — browser service | No |
| Non-English scripts | 20 languages (Browser voice) | Yes | Yes | Yes | Yes |
| Works in Firefox | Private voice + auto-scroll | Auto-scroll | Manual speed only | Recommends Chrome | App |
| Mirror / flip / full screen | Yes / yes / yes | Yes | Yes / yes / yes | Yes | Yes |
| Click a word to jump | Yes | — | — | — | — |
| Import from a transcript or SRT | Yes, one click | Paste / docs | Paste | Paste | Paste / docs |
| Paid plans | Not needed | From $15/month | Free | Free, open source | Paid plans; free plan watermarks recordings |
Checked against each product's public pages in November 2026. Speakflow and Teleprompter.com add team script libraries and recording that ScribeCue doesn't have; if you only need a prompter that keeps up with you, the free tools above are enough.
From transcript to script, and back again
Turning a talk, a podcast or an older video into a new script? Get its transcript on ScribeCue — paste a YouTube link on the home page or upload a file on Audio to text — then press Teleprompter on the result. The text opens here, without timestamps, ready to edit.
After you record, upload the take to Video to text for a transcript with speaker labels, or to Add subtitles to video to burn in captions without a watermark.
Tips for reading naturally
- Write the way you talk: short sentences, contractions, one idea per line. Blank lines become paragraph breaks.
- Keep the reading line in the top third and the column narrow, so viewers don't see your eyes travel.
- Rehearse once in Auto-scroll to find your pace; the time-left counter assumes your real words-per-minute after a few seconds.
- Use a headset or a close mic in a noisy room — the recogniser only needs to hear you, not sound good.
- Spell out numbers and names the way you'll say them ("twenty twenty-six", "Sha-NEL") so the text and your voice match.
Frequently asked questions
Is there a free voice teleprompter with no time limit?
Yes. ScribeCue's voice teleprompter is free with no sign-up, no watermark and no monthly minutes — voice scrolling works for as long as you read.
Does the voice teleprompter upload my audio?
Not in Private voice mode: the speech model runs in your browser and the audio never leaves your device. In Browser voice mode your browser sends audio to its own speech service (Google in Chrome, Apple in Safari); ScribeCue never receives it in either mode.
Which languages does it follow?
Private voice follows English. Browser voice follows 20 languages including Spanish, French, German, Portuguese, Hindi, Japanese, Korean and Chinese — pick the language under Settings.
Does it work on iPhone and Android?
Yes in Safari and Chrome on phones and tablets. Private voice needs a recent phone because the model runs on it; Browser voice and Auto-scroll work on almost any device. iPhone fills the window instead of true full screen.
Why doesn't it scroll when I start talking?
Check that the microphone is allowed in the address bar and that you are reading words from the script. In Private voice the model takes a few seconds to load the first time. If your browser has no speech recognition (Firefox), use Private voice or Auto-scroll.
Can I use it with a glass teleprompter or a mirror rig?
Yes. Turn on Mirror to flip the text left-to-right, and Flip upside down if your rig needs it. Put it in full screen on the tablet or monitor under the glass.
Can I load a script from a file or a transcript?
Import a .txt, .md, .srt or .vtt file — subtitle numbers and timestamps are removed. From any ScribeCue transcript, press Teleprompter to open its text here.
Does it record video?
No. Record with your camera, phone or screen recorder while the prompter runs. Afterwards ScribeCue can transcribe the recording and make subtitles.