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