Online Teleprompter with Voice Tracking
The FoxCue teleprompter scrolls as you speak (Cloud or on-device voice tracking) or at a fixed speed, with size, opacity, font, mirror and highlight settings.
The FoxCue teleprompter shows your script over the recorder while you record. It's never part of the recording. It can follow your voice, scrolling as you read and waiting when you pause, or scroll at a fixed speed you set. Every setting below is saved in your browser for next time.
It opens automatically when you record a script that has text. Toggle it with Teleprompter on the recorder's scene bar.
Open the teleprompter
- Recording a script: it opens automatically when the script has text.
- Recording without a script: click Teleprompter. The Teleprompter Script dialog asks What Would You Like to Say? Type or paste your text, then click Open Teleprompter.
Two ways to scroll
| Mode | How it works | Use it when |
|---|---|---|
| Voice tracking | Listens to you read and keeps the current word on the reading line. Pause and it waits. Resume and it continues. | You want to talk naturally at your own pace |
| Play (fixed speed) | Scrolls continuously at the Speed you set | You can't use the microphone for tracking, or prefer a steady pace |
Use voice tracking
Choose an engine
Pick Cloud or On-device (see the table below). The default is Cloud.
Turn on voice tracking
Click the microphone button (Voice tracking). The status shows Loading…, then Listening.
Read aloud
Start reading from the top. The word you're on is highlighted and sits on the reading line, about 40% down the panel. Words you've already said dim.
Pause and resume
Stop talking and the teleprompter waits. Ad-libs and skipped lines are fine: it finds your place again when you get back to the script.
Voice engines
| Engine | Technology | Privacy | Browsers | Notes |
|---|---|---|---|---|
| Cloud (default) | Web Speech API, English (US) | Audio is processed in the cloud by your browser's speech service | Chrome and Edge | Most accurate and instant. Needs internet. |
| On-device | Vosk | Fully private. Audio stays on your computer. | Any modern browser | Loads a speech model of about 40 MB from the FoxCue app the first time you use it in a session. Recognition then runs in your browser. |
How tracking follows you
Voice tracking compares the last 10 words you said with the script, looking from 5 words behind your place to 16 words ahead.
- It moves forward when 2 or more words in a row match, or when you say the very next word.
- If you skip ahead, it jumps to a match of 3 or more words up to 40 words ahead.
- If you stop talking, it waits.
- Text in brackets, such as cues and
[pause], isn't expected to be spoken.
The highlight only shows while voice tracking is on.
Every control
Toolbar row 1
| Control | What it does |
|---|---|
| Close (×) | Close the teleprompter. Esc also closes it. |
| Restart | Jump back to the top of the script |
| Voice tracking (mic) | Start or stop voice tracking with the selected engine |
| Play / Pause | Fixed-speed auto-scroll. Hidden while voice tracking is on. |
| Mirror | Flip the text horizontally for beam-splitter teleprompter glass. Default: off. |
| Cloud / On-device | Voice engine. Changing it stops tracking. |
Toolbar row 2
| Setting | Range | Default | Notes |
|---|---|---|---|
| Speed | 1–400 | 60 | About pixels per second. Use − / + for 0.5 steps, the slider (0.5 steps) or the number box (0.1 steps). Hidden while voice tracking is on. |
| Size | 18–64 | 30 | Text size |
| Opacity | 20–100% | 85% | Panel background opacity. Lower it to see more of your camera. |
| Font (Aa) | Serif, Sans, Bold | Serif | |
| Mark colour | Green, Amber, Cyan, White | Amber | Amber, Cyan and White stay distinct for color-blind readers. White gives maximum contrast. |
| Mark style | Fill, Underline, Bold | Fill | Fill: colored background behind the current word. Underline: a shape cue, best for color-blind readers. Bold: bold and colored. |
Moving and resizing
- Drag the toolbar by its handle to move it anywhere on the screen.
- Drag the script panel to move it. Resize it from either bottom corner: width 320–1,400 px, height 120–700 px.
- Scroll with your mouse wheel or trackpad to move through the script by hand.
- On narrow windows (under 600 px), the toolbar switches to a compact layout.
Position, size and every setting are saved in this browser.
The cue dock
If your script has recording cues, a NOW / NEXT dock appears under the script:
- NOW shows the cue you've reached, with a pulsing one-tap button: Open host for links, show slide, share screen or spotlight. Reminders have no button.
- NEXT previews the cue after that.
- Tap ✓ to mark a cue done.
- Before you reach any cue: Cues light up here as you read to them.
A [slide: N] cue also switches Full cam to the Content layout so the slide is visible.
Keyboard
| Key | Action |
|---|---|
| Esc | Close the teleprompter (press again to close the recorder) |
The studio recorder's teleprompter has no other shortcuts. For Space, speed and slide shortcuts, use the slide recorder.
Troubleshooting
| Message or problem | Fix |
|---|---|
| This engine needs Chrome/Edge (Web Speech API). Try the Offline engine. | Switch to On-device, or use Chrome or Edge |
| Mic: … | The browser blocked or lost the microphone. Allow microphone access for FoxCue and try again. |
| Offline engine: … / Couldn't load the offline model. | The on-device model failed to download or start. Check your connection and retry, or use Cloud. |
| Tracking loses my place | Read a few words exactly as written. It locks back on after 2–3 matching words. Or click Restart. |
| Text is hard to read over my camera | Raise Opacity, increase Size, or switch Mark colour to White |
| The text is backwards | Turn off Mirror |
Frequently asked questions
Does the teleprompter show up in my video?
No. The teleprompter and cue dock are only on your screen. They're never recorded.
Does voice tracking work offline?
Yes, with On-device. It downloads a model of about 40 MB once, then runs in your browser without sending audio anywhere.
Which languages does voice tracking support?
Cloud listens in English (US).
Can I use a beam-splitter teleprompter rig?
Yes. Turn on Mirror so the text reads correctly in the glass.