Letter Maps

The prompter listens for what you say and moves the text to match, which is the whole reason to use it over a timed prompter. Pause mid-sentence and the script holds still. Pick up again and it catches up without you touching a key. For anyone who's fought a runaway scroll, this is the feature that matters.

On Chrome 139 and newer, recognition can run locally instead of sending your voice to a server, which means the audio from your take doesn't have to leave the machine you're sitting at. If you're reading client scripts or anything confidential, that difference is the whole point of using this over a cloud prompter. Older browsers fall back to cloud recognition or plain auto-scroll.

Your script and settings live in the browser and aren't uploaded anywhere, so there's no account to manage and no vendor copy of your content sitting on a remote server you'll never see. The flip side is real, and it's worth saying plainly: clear your site data or move to another device and the script doesn't come with you. Keep your own copy.

When recognition isn't reliable, in a crowded room, or on a browser you don't trust with a microphone, you can switch to a fixed scroll speed and simply read to it without anything listening in. Fewer moving parts, no recognition involved. It's the mode to reach for when voice-following misbehaves on a deadline.

You tell Letter Maps which language your script is in, and voice-following tracks that language's natural pace instead of assuming a single fixed speed works everywhere. That helps if you read in something other than English, because speaking speed varies a great deal between languages.

Font size, mirroring, and a short countdown before you start are all adjustable, which sounds minor until you're bone-tired at hour three and can barely read the small print. Mirroring lets you flip the text for glass or beam-splitter rigs, and the countdown gives you a beat to settle before the take begins. Small touches, but they decide whether a session feels smooth.