YOUR NEXT CHAPTER GUIDE
Choose a voice for your characters and calls
A voice should follow the character from chat to call, with a clear indication of what can actually be spoken.
Choose from a call
Open Call and choose a voice from your library. Your custom selection is saved for this character. System voices belong to your device; a voice available on a computer may be missing on a phone.
Distinguish previews from synthesis
An audio clip can be played as a reference. On its own, it cannot speak a new sentence. Custom synthesis requires a configured engine and a compatible voice you are authorized to use. Importing does not guarantee the same voice as Character.AI.
Speak and keep the text
Calls currently use turn-taking: dictate or type, review the transcript, then send. Once a reply is saved, the selected voice can read it if synthesis is available and authorized. Dictation also depends on browser capabilities and permissions.
Understand credits
Text replies and new speech synthesis have separate allowances. System voices do not use provider synthesis credits. Replaying cached audio does not trigger new synthesis. Current voice allowances are beta settings; no top-ups are sold.