Frequently Asked Questions

💰 Cost & Licensing

Is the software free to use?

Yes! GoGospelNow is open source and completely free to use.

Component Cost
Software Free (open source)
Local models (Ollama, Whisper, Kokoro) Free
Cloud APIs (optional) Pay-per-use (minimal)
Installation help (optional) $75/hour

What are the ongoing costs?

Fully Offline (Local Models): $0/month

If you use local models for transcription (Whisper), translation (Ollama), and TTS (Kokoro), there are no ongoing costs. Everything runs on your computer.

With Cloud APIs: Minimal costs

Service Typical Monthly Cost Notes
Google Cloud TTS $0-5/month Free tier covers most churches
Groq API $0.01-0.10/month Extremely affordable
OpenAI/Mistral API $0.05-0.50/month Slightly more expensive

Bottom line: For most churches using cloud APIs, the cost is essentially negligible - less than a few dollars per month for a typical church!

Can I translate to multiple languages at the same time?

Yes! You can run multiple translations simultaneously during the same service.

Recommended Setup: For best results, use a separate computer for each language.

Languages Setup Investment
1 language 1 computer Your existing PC/laptop
2 languages 2 computers + $300-500 (mini PC)
3 languages 3 computers + $600-1000 (3 mini PCs)

Recommended Hardware:

  • Mac Mini M4 ($599+) - ⭐ Excellent choice! Unified memory acts as GPU.
  • Ryzen 7/9 + 32GB RAM ($600-900) - Matches Mac Mini performance
  • AMD Ryzen 5 mini PC ($300-400) - Good mid-range option
  • Intel N95/N100 mini PC ($150-200) - Budget option (use with cloud APIs)

📱 Mobile Listener App

How much battery does the listener app use?

The app is designed to be lightweight. Estimated battery usage: ~20-35% per hour.

Recommendation: Arrive with 50%+ battery for a 1-hour service, or 70%+ for longer services.

Tips to extend battery:

  • Lower screen brightness
  • Use wired earbuds (uses less than Bluetooth)
  • Close other apps
  • Enable Low Power Mode

How many phones can connect at once?

Congregation Size Network Needed
Up to 50 phones Standard WiFi router
50-150 phones Dual-band router or 2 access points
150-300 phones Business-grade WiFi
300+ phones Multiple access points

Each phone uses only ~30-50 kbps of bandwidth - very light.

Do phones need internet access?

No! Phones only need to connect to the same WiFi network as the translator computer. The WiFi router does NOT need an internet connection. This is perfect for:

  • Mission trips without internet
  • Remote locations
  • Buildings with poor connectivity

Why does my screen turn off during long services?

Android: Use the native Android app instead of the web page - it uses a native wake lock that keeps the screen on indefinitely.

iPhone: Go to Settings → Display & Brightness → Auto-Lock → Set to "Never" (temporarily).

🌍 Languages & Accuracy

How many languages are supported?

Mode Translation Languages TTS Voices
With Internet (APIs) 100+ languages 50+ languages
Offline (Local models) 20-30 common languages 8 languages

For less common languages, use cloud APIs for best results.

How accurate is the translation?

With base Whisper and a small local model like gemma3n:e2b, we're seeing approximately 90% accuracy.

What translates well:

  • ✅ Core sermon content and theology
  • ✅ Biblical references and quotes
  • ✅ Everyday vocabulary and phrases

What may not translate perfectly:

  • ⚠️ Pop culture references and idioms
  • ⚠️ Local slang or colloquialisms
  • ⚠️ Jokes that rely on wordplay

AI translation creates logs that can be reviewed after the fact to check accuracy.

🖥️ Server / Translator

Is the program hard to use?

No, but initial setup requires some technical knowledge.

Phase Difficulty Time
Installation Moderate 30-60 min first time
Configuration Moderate 15-30 min to optimize
Daily use Easy Just click Start!

Daily operators don't need technical skills - once it's set up, anyone can run it.

What operating systems are supported?

  • ✅ Windows 10/11
  • ✅ macOS (Intel and Apple Silicon)
  • ✅ Linux (Ubuntu, Debian, etc.)

Can the system run completely offline?

Yes! Using local models (Whisper, Ollama, Kokoro), no internet connection is required. All processing happens on the local computer.

How much computer power do I need?

Component Minimum Recommended
RAM 8 GB (API only) 16 GB+
GPU None (CPU works) NVIDIA GPU for faster Whisper
Storage 10 GB free SSD preferred

🔊 Audio Setup

How do I prevent audio feedback loops?

The translated audio should NOT be picked up by the input microphone. Options:

  1. Use a separate aux send for the translator input
  2. Use headphones/earbuds for translated audio output
  3. Keep the translator audio isolated from the main sound system

Can I use the computer's built-in microphone?

Yes, but for best results use:

  • A dedicated USB microphone close to speaker or speakers
  • A direct audio feed from the sound board into the PC

🔧 Troubleshooting

The phone can't connect to the server

  1. Make sure both devices are on the same WiFi network
  2. Check the server IP address is correct (shown in the app header)
  3. Use port 8000
  4. Check Windows Firewall isn't blocking the connection

Audio won't play on my phone

  1. Make sure your phone's volume is up
  2. Check that the phone isn't in silent/vibrate mode
  3. On iOS: Tap the ▶ play button to unlock audio
  4. Try disconnecting and reconnecting

The translation seems slow

  1. Check your computer's CPU usage and make adjustments
  2. Consider using a faster translation model
  3. Use a GPU if available for Whisper
  4. Reduce the Whisper model size (use "base" instead of "large")

Have a question not answered here?