Casola Real-time avatar & video API
Make your agents speak human
Give your agent a real-time face that listens, reacts, and stays itself across every session. One API, and your users talk to someone instead of reading text.
Not a screenshot. This is live. Talk now.
Built for conversational AI
Whether you're building support, coaching, sales, or companions, it has to see, hear, take its turn, and stay in character in real time.
-
Perception built in
the avatar listens and watches: natural turn-taking, no push-to-talk, no talking over each other.
-
Real-time by default
sub-second response so conversation feels lived, not loaded.
-
Stays in character
the same face, voice, and persona every session. Manage a whole cast of agents under one API key.
From zero to live in three steps
-
Get an API key
Sign up and grab a key from the dashboard. No sales call, no waitlist.
-
Call the API
Open a live avatar session, queue an offline render, or start a video job. Same key, OpenAI-compatible SDKs.
-
Stream or render
Stream frames in real time or poll for your finished video. Iterate fast and ship.