Getting Started

If you are opening Sona for the first time, this is the shortest route to a usable local transcription setup.

Best for

• First-time desktop installs
• Users who want the recommended offline setup instead of manual model selection
• Anyone who wants to reach Live Record as quickly as possible

Before you begin

• Sona can open normally on your machine.
• If you want the app to download the recommended model pack during onboarding, keep an internet connection available.

Install the desktop app

1. Open Sona Downloads and choose the build for your desktop system. If the site cannot load release details, use GitHub Releases as the fallback.
2. Launch the app.
3. If you are building from source instead, use the project README.

If Windows SmartScreen, macOS Gatekeeper, or Linux executable-permission prompts appear, first confirm that the installer came from Sona Downloads or GitHub Releases, then follow the operating-system prompt. Do not turn off system protection globally.

Complete First Run Setup

1. Wait for First Run Setup to appear when Sona launches.
2. Review the welcome step. Sona's recommended first success path is Microphone -> Live Record.
3. Click Continue.
4. On the models step, click Download Recommended Models. If the required models are already present, Sona can show Continue instead.
5. Wait for the recommended offline models to finish downloading and extracting.
6. Continue to the microphone step and allow microphone access.
7. If microphone permission was denied, use Try Permission Again after fixing the OS permission prompt.
8. Choose the microphone you want Sona to use for Live Record.
9. Click Start with Live Record.

What happens next

• Sona applies the recommended offline setup for local transcription.
• The app opens on Live Record.
• If setup is still incomplete, the reminder banner can reopen onboarding later.

Good follow-ups after setup

• Go straight to Live Record if you want to capture speech immediately.
• Go to Batch Import if you already have files waiting to be transcribed.
• Use FAQ and Troubleshooting if setup keeps reopening or a model still appears missing.

Helpful notes

• You can click Later during onboarding and return from the reminder banner.
• If you hide the reminder banner, it stops appearing on the home screen until setup is complete.
• You can change models later in Settings > Model Settings.
• You can change the default microphone later in Settings > Input Device.
• If model download stalls, check that your internet connection is available and the model host can be reached, then retry onboarding. If it still fails, use Settings > Model Settings to download or select models, then open Settings > General > Diagnostics for details.
• If you want command-line batch transcription later, continue to the in-site CLI Guide.