Wordtandem User Guide

Welcome

This guide covers everything in Wordtandem: getting a script in, setting up the prompter, how voice tracking behaves, recording, the desktop app, your account and plan, and what to do when something is off.

If you only read three things, read these:

Stuck with something? Use the search box at the top of this guide, or write to us and a person will reply. Support questions go to support@wordtandem.com.

Install Wordtandem

Wordtandem runs in your browser, so on most devices there is nothing to install: sign in at wordtandem.com/app and your scripts follow you.

Desktop app Mac Windows

The desktop app puts the prompter in its own window above your slides or video call, keeps the screen awake while you read, and saves recordings straight to disk. Download it here for Apple silicon Macs, Intel Macs, and Windows. It uses your Wordtandem account, and the free plan works in it.

The installers are not yet code-signed, so macOS and Windows each show a warning the first time you open the app. The download page walks through exactly what appears and what to click. It looks alarming and it is expected.

Phones and tablets

Open wordtandem.com/app in your phone's browser. The library, editor, voice tracking, and recording all work there. The Android app is prepared for release but is not in the Play Store yet.

Requirements

A close-talk microphone (a lavalier or a headset) makes every voice engine noticeably better than a laptop mic, because it hears you instead of the room.

What is Wordtandem?

Wordtandem is a teleprompter that follows your voice instead of scrolling at a fixed speed. Slow down to land a point and it waits. Stop to think and nothing moves. Go off script to answer a question and your line is still there when you come back. Nobody has to sit off camera driving the scroll, and you never have to match somebody else's speed.

It is used for recording video lessons and sermons, presenting to a camera alone, podcasts and voice-over, sales and demo videos, and anything else where you read long text on camera without a crew. It records too, with your camera behind the script so you read at the lens.

Create a script

In the library, choose New script. The editor opens with a name field, a small formatting toolbar and the text area. Type or paste your text, give it a name, and choose Save. It appears in the library straight away.

Paragraph breaks matter: the prompter keeps them, and voice tracking uses them to tell where one phrase ends and the next begins. Write the way you will speak.

Choose Add demo to drop in a short sample script if you want to try the prompter before writing your own.

Import a document

Choose Upload scripts in the library, or drag files anywhere onto the page. You can select several at once and each becomes its own script, named after the file.

From Apple Pages, Google Docs, or an old .doc

Wordtandem does not import these formats directly yet. A Google .gdoc file is only a link to the online document, while Pages and old .doc files use formats the browser's Word reader does not understand.

For now, export first: in Pages choose File › Export To › Word (or PDF); in Google Docs choose File › Download › Microsoft Word (or PDF); or save an old .doc as .docx in Word. Then upload the exported file. You can also copy the text and paste it into New script.

Managing scripts

The library lists every script with its word count. Each row has edit, export and delete. Delete asks twice so a slip can't lose a script. Export saves the plain text as a .txt file. Once you have a few scripts, a search box appears above the list.

On your other devices. When you are signed in, scripts sync to your account, on every plan including Free. Open the app on another computer or phone and sign in, and the library is there. Scripts you make while signed out stay on that device until you sign in.

Editing while prompting. Choose Edit text in the control panel to change the live script without leaving the session.

Formatting and cue points

The editor supports bold, italic, underline, two heading sizes, paragraphs and bullet lists, and a Clear button that strips formatting from the selection. Formatting shows in the prompter, so you can make a cue stand out from the words you are meant to say.

Headings become cue points

Every heading in a script is a cue point. While prompting, a section menu appears in the control panel, and the [ and ] keys (or the shoulder buttons on a game controller) jump to the previous or next one. Use headings for the sections of a talk and you can move between them without scrolling.

Stage directions

Anything in parentheses, like (pause here) or (look at camera), is treated as a stage direction. Voice tracking always skips past it, so it can neither be matched nor stall the follower. Turn on Dim stage directions under Settings › Behavior and it also reads as a note instead of script.

Main interface

Select a script in the library and choose Start prompting. Two things happen: the control panel replaces the library, and the prompter itself opens in its own window with nothing but the script in it. Put that window on a second screen, or over your camera, and keep the control panel on your main screen.

The control panel

• Listen / Pause starts and stops the session. In Auto mode the button says Play; in Manual mode it says Manual.
• Restart jumps back to the top. Edit text opens the live script in the editor. Reopen window brings the prompter window back if it was closed. Library ends the session.
• The readout shows time elapsed, your words per minute, and time left at your current pace.
• The preview pane mirrors the prompter window. Scroll it to move the prompter, click any word to jump there, or press R to resync to the current word.
• The section menu appears when the script has headings.

The prompter window

It remembers which screen it was on and what size you left it at, and opens there next time. If your browser asks to manage windows on all displays, allow it, so the window can reopen on the right monitor. If the window does not appear at all, your browser's pop-up blocker has it; allow pop-ups for wordtandem.com.

Settings

Settings live under the control panel in five tabs. Everything applies immediately, and Save as my default under Look keeps your setup for next time.

Basics

• Mode. Voice follows your speech. Auto scrolls at a fixed speed. Manual means you scroll it yourself.
• Auto speed in pixels per second, and Timed scroll: give it a target time for the whole script and the speed is worked out for you, recomputed each time you press play so it adapts to your text size.
• Text size. Bigger is easier to read from further away. ↑ and ↓ change it while prompting.
• Color scheme. Pick a saved scheme, or edit the five colors under Look and save your own.

Reading

• Layout. Line spacing, side margins, and where the reading line sits vertically (the eye-line height).
• Safe area top and bottom keeps the script clear of a phone notch, a tablet bezel, or the crop of a beam-splitter rig.
• Motion. Glide is how quickly the scroll catches up to you; Lead is how many words ahead it keeps in reserve. See Tuning the follow.
• Font. Sans is clearest at a distance; serif suits long-form reading; mono is there if you want it.

Look

• Theme colors. Text, the current-word marker, background, already-read text, and the reading line. Edit with the picker or paste a hex value. Changing any of them switches the scheme to Custom; save it under a name to keep it.
• Display. Mirror flips the text horizontally for beam-splitter glass, Flip reverses it vertically, CAPS shows everything uppercase.
• Reading aids. Dim what you have already said, highlight the current word, and draw a reading line at the eye-line.
• Tight paragraphs removes the blank space between paragraphs so more fits on screen.
• My defaults. Save the way the prompter looks right now, restore it any time, or factory-reset everything.

Behavior

• Pre-roll cue. A 3- or 5-second countdown before scrolling begins, so you can get composed after pressing record. Nobody says "action" when you film alone.
• Loop restarts from the top when Auto mode reaches the end.
• Game controller. See Remote control.
• Smooth follow glides through recognition that arrives in bursts instead of jumping. Dim stage directions is described under Formatting.

Voice & camera

• Voice recognition. Auto (VoiceFollow Pro when you have minutes, otherwise Basic), or pin one. See Basic and VoiceFollow Pro.
• Microphone. Which mic the prompter listens to. Most browsers feed Basic recognition from the system-default mic, so set your device in the operating system if the picker seems to have no effect.
• Camera. Shows your camera behind the script so you read at the lens. The browser asks permission the first time.
• Recording. See Recording.

Remote control

Everything in the prompter has a key, and anything that sends keys can drive it.

Presentation clickers and foot pedals

Most remotes and pedals present themselves as keyboards and send Page Down and Page Up, so they nudge the script forward and back with no setup. With voice tracking on, a nudge moves you and the follower carries on from there.

Game controllers

Turn on Game controller under Settings › Behavior and plug in or pair a controller. A or X plays and pauses, the D-pad nudges up and down, and the shoulder buttons jump between cue points.

Desktop shortcuts Desktop

In the desktop app the keys above work in the prompter window, and these work from any window on macOS (or from the main window on Windows). Mod is ⌘ on a Mac and Ctrl on Windows.

Recording

Under Settings › Voice & camera, pick your camera, then choose Record. The prompter records your camera and microphone at up to 1080p while you read. Choose Record again to stop, and the file is saved.

• The script is never in the file. It shows on your screen only; the recording is a clean feed with no text burned in and no watermark, on every plan, with no time limit.
• Where it goes. In the browser it lands in your downloads folder. In the desktop app you choose where to save it.
• Format. Safari usually saves MP4. Chrome, Edge, and Firefox usually save WebM. Wordtandem uses the best format your browser can record.
• Read at the lens. With the camera on, your feed sits behind the script, so your eyes stay near the lens. Mirror preview flips the feed if you find a mirrored self-view easier.

On supported desktop browsers, Wordtandem can stream a long recording directly to the file you choose. Other browsers may hold the take in memory until you stop. Check that the file saved before closing the app or tab.

Desktop appDesktop

The desktop app is the same prompter, pulled from the web on every build, plus the things a browser cannot do:

• A window that stays on top. Keep the prompter above your slides, your Zoom or Meet window, or a recording app, so the script is in front of you while you present. Toggle it from the Window menu.
• Keep-awake. Your screen will not sleep or dim while a session is running.
• Native saving. Recordings go where you choose, with a normal save dialog.
• Menus and app-wide shortcuts. Session, Window and Help menus, and the Mod+Shift shortcuts listed under Remote control. Help › Keyboard Shortcuts shows the same list inside the app.

First run

The first time you open it, a short setup walks you through signing in, choosing a microphone and camera, the library, settings, placing the prompter window, and a thirty-second practice read. Every step can be skipped, and Help › Run Setup Again brings it back.

Signing in

Sign in with your Wordtandem email and password. The app keeps a device key in the operating system's secure storage, not your password. You can see and revoke device keys from your account page, which is the thing to do if a laptop goes missing. Changing your password also signs every device out.

Updates

On a Mac the app updates itself: a new version is fetched in the background and applies the next time you launch. On Windows there is no auto-update yet; come back to the download page for new versions.

Closing

Closing the main window quits the app. Closing the prompter window only hides it; Reopen window brings it back.

Diagnostics

Help › Show Diagnostics Log opens the app's log folder, which is what to send us if something misbehaves.

Beam-splitter rigs and mirrors

On a beam-splitter rig the script is reflected in glass in front of the lens, so it has to be shown backwards. Turn on Mirror under Settings › Look (or press M), and Flip if your rig also inverts vertically. Use the Safe area sliders to keep the text inside the part of the screen the glass actually shows.

Put the prompter window on the rig's monitor and keep the control panel and preview on your own screen; the preview is never mirrored, so you can still read it.

How voice tracking works

With Mode set to Voice, choose Listen. The prompter turns your speech into text and matches it against the script, word by word, and moves the scroll to keep the word you are on at the reading line.

• Slow down and it waits. The scroll is paced to you, not to a timer.
• Stop and it stops. Silence does not move the script.
• Go off script and it holds your place. Answer a question from the room, ad-lib a story, and your line is still there when you come back to it.
• Skip ahead and it catches up. The matcher checks nearby text first. If you read from a later section, a longer, clear phrase lets it move there. A single common word cannot cause a distant jump.
• Stage directions are skipped. Text in parentheses is never matched, so a note to yourself cannot stall the follower.

Putting it back on track

If it loses you, press R to resync the scroll to the current word, click the word you are actually on in the preview, or scroll the preview to the right place. The follower carries on from there.

Basic and VoiceFollow Pro

Auto, the default, tries VoiceFollow Pro first and uses Basic if Pro is unavailable when you start. If your minutes run out during a session, Wordtandem tells you so; switch to Basic to keep voice tracking.

What counts against your minutes

Only the time VoiceFollow Pro is actually listening. Pause and the clock stops. Basic never counts. Your allowance resets on the first day of each month (UTC) and does not roll over. When it runs out, Pro stops until the reset and Basic keeps going with no limit.

Names and jargon

Before each connection, VoiceFollow Pro reads the names, capitalised terms and unusual words from the part of the script you are about to read and tells the engine to expect them. That is why proper nouns and technical vocabulary track well. It refreshes as you move through a long script.

Tuning the follow

Two sliders under Settings › Reading › Motion shape how the scroll moves, and one switch under Behavior.

• Glide is how quickly the scroll catches up to where you are. Low is the most immediate; High is smoother and lags a touch more.
• Lead is how many words ahead of the current word the follower keeps in reserve, so the next line is already in view.
• Smooth follow glides through recognition that arrives in bursts instead of jumping. Turn it off if you prefer the scroll to snap to each matched word.

If the prompter feels a beat behind you, lower Glide and raise Lead. If it feels twitchy, raise Glide. A close-talk microphone helps more than any slider.

Where your audio goes

Basic uses the speech recognition built into your browser. The audio never reaches us; how your browser handles it is covered by your browser vendor's own privacy policy.

VoiceFollow Pro sends your microphone audio to Deepgram, our speech-to-text provider, for one purpose: turning speech into text so the prompter can follow your place. Every request is opted out of Deepgram's model-improvement program, the audio is processed only for that request, and we do not keep a recording of it. The privacy policy has the full detail.

Your account

Your account page shows your plan, the VoiceFollow Pro minutes used this month, and your device keys. Account settings let you:

• Change your name, which is only used to address you in email.
• Change your email. A six-digit code goes to the new address; enter it to confirm.
• Change your password. This signs you out on every other browser and device, and revokes desktop device keys, so a lost laptop is locked out.
• Download your data as a file.
• Delete your account. This cancels any paid plan first and then removes your scripts and account. If billing cannot be reached, nothing is deleted; try again or contact us.

Forgot your password?

Use the reset link. The email you receive works once and signs out your other devices when you set the new password.

Device keysDesktop

The desktop app signs in once with your email and password and then keeps a device key, stored in the operating system's secure storage, so it never has to hold your password. Each installed copy gets its own key.

Your account page lists every active key with its device name. Revoke one and that copy is signed out the next time it talks to us. Changing your password revokes them all. You do not need a device key for the web prompter.

Plans and billing

Basic voice tracking, recording, the editor and script sync are free on every plan and never expire. Paid plans add VoiceFollow Pro minutes.

Paying yearly saves about two months. Organizations, registered 501(c)(3)s, K-12 schools and teachers: get in touch for a discount or a custom allowance with invoicing.

Changing or cancelling

Choose a plan on the pricing page; you always see the exact amount before anything is charged. Upgrades apply immediately and are prorated. Downgrades take effect at the end of the period you have paid for. Cancel any time from your account page: you keep the plan until the period ends, then move to Free and keep all your scripts. Stripe handles payment, so your card details never touch our servers.

Refunds

If it is your first payment and Wordtandem is not for you, tell us within 14 days and we will refund it. Beyond that, cancelling keeps your access until the period ends rather than refunding the remainder. The terms spell this out.

Zoom, Meet and video callsDesktop

Reading notes on a call is what the desktop app's always-on-top window is for. Open the prompter window, size it to a band just under your camera, and turn on keep above from the Window menu. Your call stays full screen underneath and the script floats over it, close to the lens, so you look at the person rather than off to one side.

Two things to know. The prompter window is a normal window, so if you share your whole screen, participants can see it; share the application window you mean instead. And with voice tracking on, your microphone is used by both the call and the prompter, which every current operating system allows.

A second screen or a tablet as the prompter

The prompter window is designed to live on a different display from the control panel. Drag it to the second screen, size it, and it will open there next time. On a rig with a tablet, open wordtandem.com/app on the tablet and sign in; your library is there. Voice tracking listens on whichever device you press Listen on, so pick the one with the better microphone.

Recording a long lesson or sermon

For a complete first recording, use the online lesson guide and free script template.

• Use headings for each section so you can jump with [ and ] if you restart a take.
• Put pauses and reminders in parentheses; they will never be matched or stall the scroll.
• Turn on the pre-roll cue so you have three seconds to settle after pressing Record.
• Check your VoiceFollow Pro minutes on the account page before a long session. If they run out while you are reading, switch the voice engine to Basic.
• Use a lavalier or headset. Room echo is the most common reason tracking lags.
• On the desktop app, keep-awake means the screen will not dim halfway through.

On a phone or tablet

Open wordtandem.com/app in Safari or Chrome and sign in. The layout adapts to the screen: the library, editor, and settings all work, and you can hide or show the prompter while you read. Upload from the phone's files app or write in the editor. The Android app is prepared for release but is not in the Play Store yet.

Troubleshooting

It doesn't hear me

Check the microphone chosen under Settings › Voice & camera, and that the browser has permission (the lock icon in the address bar). Most browsers feed Basic recognition from the system-default mic regardless of the picker, so set the device in your operating system's sound settings too.

"speech engine offline" or "reconnecting…"

Basic recognition in Chrome and Edge uses an online service; if it cannot be reached the prompter says so and stops. Check your connection, or switch the engine to VoiceFollow Pro. In Edge, also check Windows Settings › Privacy & security › Speech › Online speech recognition is on. With VoiceFollow Pro, "reconnecting…" means the connection dropped and is being retried; scrolling continues manually meanwhile and picks up again on its own.

It's behind me, or it jumps

Lower Glide and raise Lead under Reading › Motion. If it jumped somewhere wrong, press R or click the right word in the preview. Read clearly for a few words after resuming from an ad-lib so it can find you.

The prompter window never opened

Your browser blocked a pop-up. Allow pop-ups for wordtandem.com and choose Reopen window. If it opens on the wrong screen, drag it once; the position is remembered. Allow "manage windows on all displays" if the browser asks.

No camera in the list

Choose Allow camera access, then accept the browser's prompt. If another app has the camera open, close it first.

"Recording is not supported in this browser"

Make sure the browser has permission to use both the camera and microphone, then reload the page. If it still fails, try the current version of Chrome, Edge, or Safari.

My scripts aren't on my other device

Sign in on both devices with the same email. Scripts made while signed out stay on that device until you sign in there.

The desktop app says it can't be opened

That is the unsigned-app warning. The download page shows the exact steps for macOS (System Settings › Privacy & Security › Open Anyway) and Windows (More info › Run anyway).

Frequently asked questions

Does the free plan expire?

No. Free is free, with no card and no clock. It includes 15 minutes of VoiceFollow Pro a month and unlimited Basic.

Is there a watermark?

Never, on any plan. A free recording is the same clean 1080p file a paying customer gets.

Can two people share one account?

One sign-in belongs to one person. If several people in an organization use Wordtandem, each needs their own; talk to us about a team arrangement.

Does it work offline?

The prompter, the editor and Basic tracking in a browser that does recognition locally can work without a connection; Basic in Chrome and Edge, and VoiceFollow Pro, need one.

Which languages?

English today.

Can it read a Google Doc directly?

Not yet. Download it as Word or PDF and upload that, or paste the text.

How does it compare to other teleprompters?

We wrote honest, row-by-row comparisons with Speakflow, Teleprompter.com and ShareSpeak, including the rows where they win.

Contact us

Something not covered here, or not working the way this guide says it should? Write to us and a person will reply, usually within a day or two. For support, email support@wordtandem.com. If it is a bug, the diagnostics from Help › Show Diagnostics Log (desktop) or the browser and operating system you are on save a round trip.