The teleprompter that follows your voice — even when you go off script.
By Thiago Lourenço Martins
Sayscroll is a teleprompter that runs right in your browser and uses voice recognition to follow what you say, word by word. Forget manually adjusting scroll speed like on old-school teleprompters: you paste the text, hit a button, and start talking — words fade out as you say them, and the text scrolls on its own to keep the current line centered. If you stumble, skip a line, or decide to improvise, it simply waits — and finds the exact spot again the moment you return to the script.
One thing worth clearing up right away: it doesn't write your script or build any slides for you. You still do the writing — its only job is making sure the text keeps up with your pace once it's showtime.
You memorized the intro, hit record, and lost your place halfway through the second sentence. Cut. Start over. Again. A 40-second script turns into half an hour of recording because every time your eyes drift to the paper — or the text scrolls too fast on screen — you lose the thread.
That's the voice-tracking accuracy Sayscroll itself advertises, across more than 60 languages — Portuguese included. You talk at your normal pace, it keeps up; you hesitate, it pauses; you repeat a line to make a point land, it follows along again, with no input from you.
Records videos for YouTube or Reels and lives the classic dilemma: memorize the script and stumble, or read it and lose eye contact with the camera. Pastes the text into Sayscroll, looks at the lens, talks normally — the text keeps up.
Has to record a lesson with a memorized intro full of technical terms that can't be flubbed. Uploads the script, opens presentation mode, and keeps eye contact with the camera instead of glancing down at paper.
Has an opening and closing script ready, but wants to improvise answering live chat questions in the middle. Sayscroll waits while you go off script and picks up on its own when you come back.
Click "Try it below" to jump straight to the test box — no account needed to try it.
The free demo accepts up to 600 characters, roughly a short video intro.
longer scripts require a paid planThe screen switches to presentation mode and asks for microphone access on your device.
Every word you say fades out of the text, and the current line stays centered on screen, at your own pace.
Skip a line, repeat a section, or pause to think: Sayscroll waits and finds the right spot on its own when you return to reading.
Text size, column width, black or white background, mirroring (for physical teleprompter glass) and fullscreen all live in the presentation screen's own controls.
Hey, how's it going? I'm [your name] and today I'm going to show you how this works in practice. If you've made it this far, you've probably already gotten tired of recording the same video over and over just because you lost your place in the script mid-sentence. So take a breath, relax — and let's get straight to it: in three simple steps, you'll record your first clean take without stumbling once.
Paste this text into Sayscroll's demo box, hit "Hit Present", and talk at your normal pace. Then try pausing on purpose mid-sentence — the point is to watch the text wait for you.
Write a short 30-second script — the opening of a video you need to record this week. Paste it into Sayscroll's demo box, hit "Hit Present", and talk normally. Halfway through the second sentence, deliberately jump to another part of the text.
It worked if the text catches up with your speech again within a few seconds, with no clicking needed to "put the script back in place."
I tested it by pasting a video intro into the demo box, and the voice recognition genuinely works: when I deliberately backtracked a sentence just to check, it caught up on its own, no stumbling. If you re-record videos five times just because you lose your place in the script, that alone is worth the try. The catch is the free tier is too short for anything beyond a taste — 600 characters won't even fit a longer intro — and everything past that is priced only in dollars, billed by monthly usage hours. If you record one video a week, no problem. If you're on camera all day, the Pro plan's 100 hours for $50/month might not be enough.
Checked against official sources on July 27, 2026.
It's free to test, that's it. The demo on the homepage accepts up to 600 characters, no account. Unlimited script length and real usage hours only come with paid plans — and only in dollars, since the tool has no pricing in reais.
$0 — demo on the homepage, scripts up to 600 characters, no account or card.
$10/month (first month $5) — 10 hours of use per month, unlimited script length.
$20/month (first month $10) — 30 hours of use per month. For frequent recorders.
$50/month (first month $25) — 100 hours of use per month. For studio pros live all day.
Important detail: only time with the microphone actually active counts as "usage hours" — it pauses on its own after 15 seconds of silence. Billing runs through payment partner Polar, cancellable anytime. Official prices checked in July 2026, always in US dollars.
There's a free demo right on the homepage, no account needed: paste a script up to 600 characters and hit "Hit Present". For unlimited script length and real monthly usage hours, you need a paid plan, starting at $10/month.
Not for the free demo. A card is only needed if you subscribe to a paid plan, billed through the payment partner Polar, cancellable anytime.
Yes. Portuguese is among the 60+ supported languages, and you can even mix Portuguese with another language in the same script.
It runs straight in your phone's browser, and you can add it to your iPhone or Android home screen to open in fullscreen with one tap, like an app. There's no version to download from the App Store or Play Store.
The microphone listens to what you say in real time and matches it word by word against the pasted text. Each spoken word fades out, and the text scrolls to keep the current word centered on screen.
It waits. When you come back to the text — even jumping to another section or repeating a line — Sayscroll finds the right spot on its own, with no input from you.
Only the time the microphone is actively listening counts. It pauses automatically after 15 seconds of silence and stops when you reach the end of the script or exit presentation mode.
According to the official site, your script stays on your device and only the audio is streamed in real time for transcription — not a fully offline process, but the company states it doesn't store that recording.
If what's missing is the SLIDE itself — you don't even have a presentation yet — Gamma builds the design from a text prompt. Sayscroll comes in later, when it's time to talk.
Another way to turn a PDF or text into a ready-made presentation, before you even think about the speaking script.
If the plan is to not appear speaking at all — just narrate over the video — swap the teleprompter for an AI-made narration instead.
Every day, a new tool explained.
Get it on WhatsApp