Voice-Activated Teleprompter

Teleprompter with voice tracking for live presentations. Choose between browser-based speech recognition (easy setup) or offline Vosk (privacy-focused).

Screenshots

Settings

Prompter Mode

Quick Start

Browser mode (no setup):

1. Open frontend/index.html in Chrome/Edge
2. Paste your script, click Start
3. Speak - words highlight as you read

Vosk mode (offline/private):

./deploy.sh

Then access via your server URL.

Features

Speech Engines

• Browser - Uses Web Speech API, works in Chrome/Edge, 13+ languages
• Vosk - Offline recognition, no data leaves your server
• Unavailable options are automatically disabled

Pacing Controls

• Hold Space to advance words manually
• Double-tap Space to jump to next line
• Ctrl to pause/resume
• ↑/↓ to adjust speed (CPM-based with punctuation pauses)
• Number keys to jump to quick-nav cue points (auto-pauses)
• Press ? for keyboard shortcuts

Dialect Support

• Match Tolerance slider (50-100%)
• Lower values enable fuzzy matching for dialects (Swiss German, etc.)

Script Markup

Regular text is matched by voice.

*emphasis* renders yellow/underlined
**strong** renders red/bold (multi-word spans work)

[stage direction] shown but not voice-matched

[1] quick-nav cue point (jump with number keys)
[15] multi-digit quick-nav points work too

{talking points
- click to advance
- bullet list}

Display Options

• Adjustable font size and scroll position
• Word highlight toggle (red box or subtle underline)
• Horizontal/vertical mirroring
• Auto-fullscreen when presentation starts
• Settings persist across page reloads

PWA Support

• Installable on mobile and desktop
• Works without browser chrome when installed
• Caches assets for faster loading

File Management

• Open .md - Load script from markdown file
• Save .md - Export script to markdown
• Save/Load - Cross-device sync via 5-char code or QR

Manual Setup (Vosk)

cd /opt/teleprompter/backend
python3 -m venv venv && source venv/bin/activate
pip install -r requirements.txt

# Download a model from https://alphacephei.com/vosk/models
mkdir -p models && cd models
wget https://alphacephei.com/vosk/models/vosk-model-small-de-0.15.zip
unzip vosk-model-small-de-0.15.zip

uvicorn main:app --host 0.0.0.0 --port 8000

Keyboard Shortcuts

Key	Action
Space (hold)	Advance words
Space Space	Jump to next line
Ctrl	Pause / Resume
↑ ↓	Adjust speed
0-9	Jump to quick-nav
Esc	Stop presentation
?	Show help

Troubleshooting

• Mic not working? Browser mode needs HTTPS in production. Check permissions.
• Vosk not connecting? Ensure backend is running, check WebSocket proxy config.
• Test page: /static/test.html shows mic levels and device info.

Reverse Proxy (Vosk mode)

Needs WebSocket support. For Caddy:

teleprompter.example.com {
    reverse_proxy localhost:8000
}

With authentication: PWA files must bypass auth for install to work:

teleprompter.example.com {
    @pwa path /static/manifest.json /static/sw.js /static/icon-*.png
    handle @pwa {
        reverse_proxy localhost:8000
    }

    handle {
        # your auth config here
        reverse_proxy localhost:8000
    }
}

License

MIT