← All guides

Microphone and speech models

Set up local transcription and choose an appropriate speech model.

Check the audio path

  1. Confirm Windows can hear the chosen recording device and that the service audio reaches it.
  2. Select the device in the dashboard, adjust gain, and save changed preferences. Watch the meter while testing speech.
  3. If text is missing, check the device, Windows microphone permissions, mute controls and the level at the source.

Choose a model

  1. AI transcription requires optional activation. Activate on the website, sign into VerseCue, then open Preferences → Speech. VerseCue Clarity is bundled and is the English-focused default.
  2. Choose an optional optional model from the dropdown to download it. Internet is needed for the download.
  3. Click Save to activate your prepared choice. Changing from VerseCue Clarity may reduce transcription or Scripture-detection accuracy; test with your own audio.
  4. Manage downloaded models lets you remove inactive optional models. Active models, pending selected models and the bundled default are protected from deletion.