Voice replies and recorded messages
Text-to-speech reads a reply aloud. Voice messaging records your message instead of typing. Neither is the same as an uninterrupted live call.
Listening to a reply
A creator can assign a voice, or you can choose an available voice for a chat. The reply's playback control generates audio from its text. An existing message can be played again.
Changing the voice can regenerate an earlier reply in that voice. The written reply and its spoken delivery are separate choices.
Sending a voice message
Press and hold the recording control to speak, then review the transcript before sending. Send modes include audio with text, audio only and text only.
Check names and unusual terms in the transcript. If the microphone does not record, review the browser or device permission before trying again.
Custom voices
Create Voice accepts an audio sample you have the rights to use. After creating it, choose it from My Voices for spoken playback. Follow the sample and setup steps before recording a longer file.
Voice slots and account access are separate from the number of written messages. Review the available capacity before creating another voice.
Access and recordings
Check the controls available in your account. A free text-model allowance does not establish which voice options are included.
Use privacy precautions for recordings as well as prompts. A recording, its transcript and the character's reply are different pieces of information.
Frequently asked questions
Can I hear a character's text reply?
Text-to-speech provides spoken playback once a voice is selected for the character or chat.
Can I review a voice message before sending?
Yes. Review or edit the transcript before sending, then choose the available audio or text send mode.