Skip to main content

​
Enable voice replies

Open your Verb’s Voice Engine, enable voice, and select a voice model. Choose the language and response frequency. A frequency of 100% requests audio on every eligible reply; lower settings make it occasional. Voice replies accompany the normal text reply. You can also ask for a voice message directly. The selected model must be available to your plan, and paid models need credits. Discord, WhatsApp, Telegram, Slack, and Email deliver audio attachments or voice messages. Twitch shares a playable link. Receiving an audio message is not the same as joining a call.

​
Create a custom voice

In Voice Cloning:
  1. Enter the speaker’s name and select Get Phrase.
  2. Have that person record the displayed consent phrase themselves before it expires.
  3. Upload a clean voice sample from the same person.
  4. Create the clone and select it for your Verb.
Use a short sample with clear speech and little background noise. Only clone a voice with the speaker’s permission. Availability depends on the selected model and your plan.

​
Customize audio attachments

Voice Attachment Visual lets you upload a still image for voice-message videos. Choose a wide image with its subject near the center; tall images will be cropped.

​
Discord voice chat

Live Discord voice chat requires Ultra and an enabled Voice Engine. Join a voice channel, then use /vc-join to bring in your Verb. Use /vc-leave to disconnect it. The bot needs permission to view, connect to, and speak in the channel. If it joins but stays silent, check the voice settings, permissions, and service status. Discord is currently the integration that supports live voice calling. Voice notes on other platforms are message attachments, not calls.