Documentation
Everything you need to evaluate, deploy, and integrate.
Written for two readers: the person deciding whether AIvatars fits, and the engineer wiring it into a product. Documentation is in English.
Start here
What AIvatars is
A real-time AI avatar that listens, understands, and answers out loud — on a website, in an app, or on a kiosk.
How it works
The path from a spoken sentence to a talking face, and why it lands in about two seconds.
Avatars and voices
2D photoreal versus 3D Unreal Engine characters, how a custom avatar is made, and how voices are chosen.
Deploy
Deployment shapes
Hosted SaaS, on-prem in your own cloud account, or an appliance on your LAN — the same product, three delivery models.
Hardware requirements
What GPU each avatar type needs, the recommended kiosk spec that runs the whole stack on one machine, and network requirements.
Getting started
What a pilot looks like week by week, what we need from you, and how success is judged.
Security and data
Isolation model, what leaves your network, how devices authenticate, and what we can hand your security reviewer.
Integrate
Embedding an avatar
The fastest integration — drop a hosted avatar page into your site or app with an iframe.
API reference
Session lifecycle endpoints — create, inspect, interrupt, and end a conversation.
Realtime protocol
The WebSocket that carries audio and conversation events, and how the avatar video arrives.
Knowledge and conversation flows
Where the avatar's answers come from, how structured procedures are defined, and how it calls your systems.