Whisperly User Guide

Whisperly is a fully local, cross-platform desktop application that converts audio files to text using advanced AI technology. All audio processing is done locally on your computer, ensuring privacy and data security.

• Privacy First: All processing is local, your audio never leaves your device
• Multi-format Support: MP3, WAV, FLAC, M4A, AAC, OGG, WMA
• High Quality: Powered by OpenAI Whisper model
• Cross-platform: Windows, macOS, Linux
• One-time Purchase: Lifetime license, no subscription
• Offline Operation: No internet required after initial setup

1. Download Whisperly from the official website
2. Run the installer and follow the setup wizard
3. On first launch, the AI model (~1GB) will be downloaded
4. After download, you can use Whisperly fully offline

*Internet is only required for the initial model download.

• Welcome Screen: Large icon, "Select Audio File" button, drag & drop support
• Workspace: Left - file info/progress/export; Right - transcription results and editor
• Status Indicators: Local processing, Ready, Processing, Loading, etc.

1. Select audio file (multiple formats supported)
2. Automatic format detection and audio analysis
3. Preprocessing (convert to 16kHz mono)
4. AI transcription to generate text
5. Post-processing (punctuation, timestamps, formatting)

• Supports auto language detection, quality/speed settings, CPU/GPU selection

• Model Selection: Tiny/Base/Small/Medium/Large (speed vs. accuracy)
• Compute Device: Auto/CPU/NVIDIA GPU/Apple Metal
• Output Format: TXT/SRT/JSON/CSV
• Output Directory: Custom save path

Click "Save Settings" after changes. Some settings require restart.

• Export as TXT, SRT, JSON, CSV
• One-click copy transcription to clipboard

1. After transcription, click "Export"
2. Select format and save location

• Model download failed: Check network and disk space
• Slow processing: Close other apps, try smaller model or GPU acceleration
• Unsupported audio: Convert to MP3/WAV and ensure file is not corrupted
• Crash: Update version, check system requirements, try CPU-only mode

Optimization: Prefer GPU, close unnecessary apps, choose suitable model, ensure good cooling.

• Ctrl+O: Open audio file
• Ctrl+S: Save settings
• Ctrl+Q: Quit application
• F1: Open help
• Ctrl+C: Copy transcription
• Ctrl+E: Export transcription
• Ctrl+,: Open settings
• Enter: Save settings
• Escape: Cancel/close panel

• Batch processing: Multiple files sequentially, batch export
• Audio analysis: RMS, zero-crossing rate, spectral centroid, SNR
• Multi-language: Auto/manual selection, 99+ languages, mixed content
• Quality metrics: Confidence score for each segment
• Custom dictionary: Add specialized terms for better accuracy

• Version and license info in settings
• Auto update check when online, free lifetime upgrades
• For help, visit the official website or contact support