Performance Tips

Optimize Speakly for your system

Local Model Performance

  • Use smaller models (base, small) for faster results
  • Enable GPU acceleration if available
  • Close other heavy applications during transcription
  • Consider BYOK for faster processing on slower hardware

Memory Usage

  • Larger models require more RAM
  • large-v3 needs ~4GB RAM
  • Close unused tabs/applications if memory is limited
Documentation - Speakly