NotchPrompt v2.0: word tracking, floating mode, and 30+ new features
tl;dr
v1 was a teleprompter that hides in your Mac notch and stays invisible on screen shares.
v2 is that, plus real-time word tracking, floating and fullscreen modes, multi-page scripts, PPTX import, OpenDyslexic, macOS Services, a URL scheme, and 30+ other things.
the big one: word-level real-time highlighting
as you speak, each word lights up and dims as you pass it.
not paragraph-level. not sentence-level. word by word, in real time.
it's powered by a fuzzy matching algorithm that handles accents, speech recognition errors, and fast speech. if you say "gonna" instead of "going to", it still tracks. if you skip a word, it catches up. tap any word to jump the reading position there manually.
this is the feature that changes how the whole app feels. you stop reading at a teleprompter and start having a conversation with your script.
3 overlay modes
Notch , the original. expands from the Mac notch with a Dynamic Island animation. anchored below the camera, invisible on screen share.
Floating , a draggable, always-on-top window. put it anywhere on screen. useful when you're not on a MacBook with a notch, or when you want the script next to a specific window.
Fullscreen , takes over an entire display. for dedicated teleprompter setups: a second monitor, a mirror rig, a recording studio. the script fills the screen.
3 listening modes
Word Tracking , per-word highlight in real time. the main mode. uses the microphone to follow exactly where you are in the script.
Voice-Activated , scrolls while you speak, pauses when you stop. no word-level tracking, but great for natural pacing without thinking about speed.
Classic , constant WPM scroll, no microphone needed. the traditional teleprompter mode. set your speed and go.
multi-page scripts
long scripts are now split into pages with sidebar navigation.
each page shows word count, character count, and estimated read time. a green badge marks pages you've already read through.
PPTX import , drag a PowerPoint file onto the app. it pulls the presenter notes from each slide and creates one page per slide. if you use Keynote, there's a warning with exact export instructions.
auto-advance , when you finish a page, a configurable countdown (1–10 seconds) automatically moves to the next one. no tapping required.
file format , scripts save as .notchprompt files. open them later, share them, keep them in a folder.
templates , four built-in starting points: Sales Pitch, Presentation, Interview, and Blank.
dictation mode , dictate text directly into the editor with your voice. write your script without typing.
overlay controls
everything you need without leaving the overlay:
- resizable height , drag the handle to expand or shrink the overlay
- in-overlay page picker , tap the page indicator to jump to any page without opening the sidebar
- hover-to-pause , hover over the overlay to pause scrolling
- scroll wheel , scroll the script manually with your trackpad or mouse wheel
- last spoken words , shows the last few words the speech recognizer heard
- audio waveform + progress bar , visual feedback that the mic is active
- elapsed time timer , MM:SS counter so you know how long you've been speaking
appearance
font families: Sans (default), Serif, Mono, and Dyslexia (OpenDyslexic). the OpenDyslexic option was the most requested feature after v1.
font sizes: XS, SM, LG, XL.
text colors: White, Yellow, Cyan, Green, Lavender.
cue annotations , wrap text in [brackets] to mark it as a cue. separate color and brightness control for cue text so it stands out from the script.
blur background , frosted glass effect with adjustable opacity. works in all three overlay modes.
system integration
macOS Services , right-click any text anywhere on your Mac → "Read in NotchPrompt". the text opens directly in the app. works in Safari, Notes, Mail, anywhere.
URL scheme , notchprompt://read?text=.... automate it, link to it, trigger it from Shortcuts or any other app.
multi-display support , the overlay follows your mouse to any connected screen.
auto-update checker , silent check on launch. no popups unless there's actually an update.
speech settings
9 languages: English (US), English (UK), Spanish, French, German, Italian, Portuguese, Japanese, Chinese.
microphone selection , pick any CoreAudio input device. if you have a dedicated mic, a USB interface, or AirPods, you can select it directly.
keyboard shortcuts
| shortcut | action |
|---|---|
| `⌃ + `` | toggle overlay |
⌃ + Space | start / stop scroll |
⇧⌘O | toggle overlay (global) |
⇧⌘S | start / stop scroll (global) |
⇧⌘R | reset position (global) |
⇧⌘] | next script |
⇧⌘[ | previous script |
everything runs on-device
no cloud. no account. no subscription.
speech recognition uses Apple's on-device APIs. scripts stay on your machine. nothing leaves the app.
what's next
v2.1 roadmap:
- browser remote viewer , watch the teleprompter on your phone over local WiFi. no app install needed, just open a URL.
- director mode , a remote operator scans a QR code and sends the script to the presenter's machine. useful for live events and broadcast setups.
- external display / mirror mode , fullscreen on a second monitor with optional horizontal flip for mirror teleprompter rigs.
- AI script polish , on-device suggestions to improve pacing and readability. no API keys, no cloud.
built by
Ali Arain , aliarain.com · github.com/aliarain
RaptrX , raptrx.com