Voice & AvatarFX: characters you can hear and see.
Two features moved Character.AI beyond text: voice calls that feel like phoning your character, and AvatarFX — give it one image and it produces video of that face talking, emoting, even singing. Here’s how to use both well.
01Voice: call your character
Any character can talk. Pick from a library of voices (or characters ship with one chosen by their creator), hit the call button, and speak naturally — it listens and answers in real time. Language practice is the killer use: speaking Spanish to a patient tutor beats typing it by a mile. Free tier meters calls; c.ai+ makes them unlimited.
A language tutor or interview coach — voice turns rehearsal into the real thing. Ten minutes out loud is worth an hour of typed practice.
02AvatarFX: one image, living video
AvatarFX takes a single image — portrait, painting, pet, cartoon — plus audio or a script, and generates video of it speaking with matched lip-sync, expressions, and gestures. It handles multiple speakers and holds consistency across longer clips. The uncanny power comes with a rule: only animate images you have the right to use — your own creations, not photos of real people without their consent.
03Scenes, Streams, and the feed
Scenes drop your character into a pre-built cinematic moment; Streams let you pick any two characters and a topic, hit play, and watch them interact. Creations can be shared to the community feed — which is also the best place to see what skilled creators are doing before you build your own.