Cloud Transcription

Cloud transcription sends your audio to Groq for processing using the Whisper Large V3 Turbo model. It supports more languages and is generally more accurate than local models.

Vorn Cloud (proxy mode)

This is the simplest option. Vorn Voice sends audio through the Vorn proxy server to Groq. It is included with your subscription at no additional cost.

To use it:

1. Open settings and go to the model section.
2. Select Cloud as your transcription model.
3. Make sure you are signed in with an active subscription or beta access.

Usage is rate-limited per account to keep the service available for everyone.

BYOK (Bring Your Own Key)

If you have your own Groq account, you can use your API key directly. This bypasses the Vorn proxy and gives you direct access.

1. Get an API key from Groq. Keys start with gsk_.
2. Open settings and go to the cloud section.
3. Enter your Groq API key. It is stored locally and masked in the UI.
4. Select the cloud model.

To remove your key later, use the clear button in settings.

Language selection

Cloud transcription supports many languages. The default is English (en). You can change the language in the cloud settings section. Set the language code (e.g., es for Spanish, fr for French, de for German) and it will be used for all cloud transcriptions.

When to use cloud

• You need to dictate in a language other than English.
• You want higher accuracy than local models provide.
• You are on a reliable internet connection.

For English-only dictation with a good microphone, local models (especially Vorn Standard or Vorn Pro) are usually fast enough and keep your audio on device.